STUDIES OF THE FLUIDIZED LIME-BED
COAL COMBUSTION DESULFURIZATION SYSTEM
Part 1 - Design of the High Pressure
Fluidized Bed Combustion Lime
Regeneration Pilot Unit - The FBCR Miniplant
Part 2 - Factors Affecting NOy Formation and
Control in Fluidized Bed Combustion
By
A. Skopp
M.S. Nutkis
G. A. Mammons
R. R. Bertrand
FINAL REPORT
JANUARY 1, 1971 - DECEMBER 31, 1971
Prepared Under Contract CPA 70-19
for the
Control Systems Division
Office of Air Programs
U.S. Environmental Protection Agency
ESSO RESEARCH AND ENGINEERING COMPANY
Government Research Division
Linden, New Jersey
GRU.13GFGS.71
-------
FOREWORD
SUMMARY
TABLE OF CONTENTS
Page
1
8
INTRODUCTION
12
PART I - DESIGN OF THE HIGH PRESSURE FLUIDIZED BED COMBUSTION
LIME REGENERATION PILOT UNIT - THE FBCR MINIPLANT
1.
13
DESIGN AND OPERATING PARAMETERS
2.
THE FBCR MINIPLANT DESIGN
17
2.1 Combustion and Regeneration Reactors
2.2 Bed Support and G~s Distribution Grids
2.3 Cyclones and Discharge System
2.4 Combustor Heat Removal
2.5 Coal and Limestone Preparation and Injection
2.6 Solids Transfer
2.7 Bed Level Control
2.8 Reducing Gas Generator
2.9 Supporting Structure
2.10 Miscellaneous Equipment
20
25
25
31
33
35
37
38
38
41
3.
DESIGN VERIFICATION WITH THE COLD MODEL TEST UNIT
42
3.1
3.2
3.3
Features of the CMTU
Solids Transfer Studies
Solids Fluidization and Entrainment
42
44
48
3.3.1
3.3.2
Predicted Regenerator Bed Slugging Height
predicted Combustor Bed Slugging Height
48
50
PART II - FACTORS AFFECTING NOx FORMATION AND CONTROL IN
FLUIDIZED BED COMBUSTION
54
1.
EXPERIMENTAL APPARATUS, MATERIALS AND PROCEDURES
55
1.1
Fluidized Bed Coal Combustion Unit
55
1.1.1
1.1.2
1.1.3
Coal Feeding Equipment
Fluidized Bed Reactor
Gas Cleanup and Analysis System
55
55
58
1.2
Fixed Bed Reactors
60
1. 2.1
1.2.2
2.5 Inch Reactor System
1 Inch Reactor System
60
62
-------
TABLE OF CONTENTS (CONTINUED)
Page
1.3
Feed Materials
62
1.3.1
1.3.2
1.3.3
1.4
Limestone
Coal
Alundum and CaS04
62
64
64
Experimental Procedures
65
1.4.1
1.4.2
2.
Fluidized Bed Combustor
Fixed Bed Reactors
65
65
GENERAL CONSIDERATIONS OF NOx FORMATION AND CONTROL
IN FLUIDIZED BED COAL COMBUSTION
66
3. EXPERIMENTAL RESULTS
3.1 Effec t of Bed Temperature and Excess Air
on NO Emiss ions
x
3.1.1 Bed Temperature
3.1.2 Excess Air
3.2 NO-CO Reaction Studies
3.2.1
3.2.2
3.3
69
69
70
74
79
Effect of Sub-Stoichiometric Combustion
Effect of H20 on the NO-CO Reaction
82
. 84
87
NO-S02-CaO Reaction System
88
3.3.1
3.3.2
3.3.3
3.3.4
3.4
Examination of Rate Controlling Mechanism
in the NO-S02-CaO Reaction System
Apparent Reaction Order with Respect
Temperature Dependence of NO-S02-CaO
Miscellaneous NO-S02-CO-CaO Reaction
90
93
94
to NO
System
Stud ies
Examination of Two Stage Combustor Operation for
NO Emissions Control
x
99
REFERENCES
NOMENCLATURE
102
103
APPENDIX 1 - PREDICTED S02 CONCENTRATION FOR PRESSURIZED REGENERATION 104
APPENDIX 2 - CALCINING PROCEDURE
105
N-1359 106
107
108
APPENDIX 3A - MINIMUM FLUIDIZING VELOCITY - 20% SULFATED LIME
APPENDIX 3B - MINIMUM FLUIDIZING VELOCITY - ALUNDUM
APPENDIX 4 - SUMMARY OF OPERATIONS - ALUNDUM AND CaS04 BEDS
-------
TABLE OF CONTENTS (CONTINUED)
APPENDIX 5 - SAMPLE CALCULATION OF FRACTIONAL CONVERSIONS TO NO
APPENDIX 6 - SUMMARY OF STUDIES OF NO - S02 REACTION SYSTEM
APPENDIX 7 - METHOD OF SULFATING LIME N-1359 FOR USE IN FIXED BED
REACTOR STUDIES
Page
112
113
118
-------
No.
1-1
1-2
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
LIST OF TABLES
Design Basis for the FBCR Miniplant
Parameters in CMTU Entrainment and Fluidization Studies
Composition of Limestone Used in Esso FBC Program
Composition of Coal Used in Esso FBC Program
Differential Reactor Results of NO-CO-CaS04 Reaction System
Differential Reactor Results of NO-CO-CaO Reaction System
Results of Study of NO-CO-H20 Reaction System
Effect of Bed Material on the NO-S02 Reaction
Conversion of NO at Constant W/F Indicates Film Diffusion
Not Controlling
Additional Fixed Bed NO-S02-CaO Reaction Studies
NO-CO-S02-CaO Reaction Studies
Page
13
48
62
64
79
80
85
89
90
94
98
-------
No.
1-1
1-2
1-3
1-4
1-5
1-6
1-7
1-8
1-9
1-10
1-11
1-12
1-13
1-14
1-15
1-16
1-17
1-18
1-19
1-20
1-21
1-22
1-23
LIST OF FIGURES
Page
1
FBCR Miniplant Flow plan
3
2
FBCR Miniplant Assembly
4
Relationship Between Regenerator Diameter and
Operating Velocity
15
Material Balance for the FBCR Miniplant
16
FBCR Miniplant Flow plan
18
FBCR Miniplant Assembly
19
Combustor Shell
21
Refractory Lined Combustor
22
Regenerator Shell
23
Refractory Lined Regenerator
24
Combustor Air Distributor plate
26
Combustor Cyclone
27
Regenerator Cyclone
28
Receiver Vessel #1
29
Receiver Vessel #2
30
Combustor Cooling Coils
32
Coal/Limestone Feeding System
34
Pulse Solids Transfer Pot
36
Supporting Structure
39
Vessel Support Legs
40
The Cold Model Test Unit
43
Solids Transfer Reservoir
46
Simulated Heat Transfer Loop
47
Mean Slug Height Observed in CMTU Studies
51
Maximum Slug Height Observed in CMTU Studies
52
-------
No.
1-24
2-1
2-2
2-3
2-4
2-5
2-6
2-7
2-8
2-9
2-10
2-11
2-12
2-13
2-14
2-15
2-16
2-17
2-18.
2-19
2-20
LIST OF FIGURES (CONTINUED)
Particle Size Distribution of Bed Material Used in
CMTU Studies
Esso Fluidized Bed Combustion Unit
Fluidized Bed Combustor
Staged Combustion Reactor
Fixed Bed Reactor
Particle Size Distributions of Limestone and Coal Feeds
to Esso FBC
Thermodynamic Equilibrium NO Concentration
NO Emissions Using Different Bed Materials
NO Emissions as a Function of Bed Temperature
Effect of Bed Temperature on Conversion of Inlet N to NO
(Alundum Bed)
Effect of Bed Temperature on Inlet N Conversion to NO
(CaS04 Bed)
Effect of 02 in Flue Gas on NO Emissions (Alundum Bed)
Effect of 02 in Flue Gas on NO Emissions (CaS04 Bed)
Effect of 02 in Flue Gas on N Conversion to NO (CaS04 Bed)
Effect of Temperature on the NO-CO Reaction Rate
NO Emissions as a Function of Percentage Stoichiometric Air
Typical Emissions with Limestone Bed
Apparent Reaction Order with Respect to NO of NO -SO -CaO
x x
Reaction System
Temperature Dependence of NO-S02-CaO Reaction System
S02 Equilibrium Over CaS03
Staged FBC Results
Page
53
56
57
59
61
63
67
71
72
75
75
76
76
78
81
83
87
92
95
96
101
-------
FOREWORD
This report was prepared by the Esso Research and Engineering
Company, Linden, N.J. for the Control Systems Division, Office of Air
Programs of the U.S. Environmental Protection Agency (EPA) under Contract
CPA 70-19.
The work described in this report was performed over the period
January 1, 1971 to December 31, 1971.
Mr. S. L. Rakes was the Project
Officer for EPA.
The authors wish to express their appreciation to Messrs. H. R.
Si1akowski and W. H. Reilly for performing the laboratory work described
in this report, and to Mr. R. K. Bryant for preparing the detailed engineer-
ing drawings of the fluidized bed pilot unit designed as part of this program.
-------
- 1 -
SUMMARY
Esso Research and Engineering Company has carried out a two phase
program for the Office of Air Programs of the u.S. Environmental Protection
Agency aimed at furthering the development of fluidized bed combustion of
coal as a new, low-polluting boiler technique for use in electric power
plants.
In fluidized bed combustion, the coal is burned in a hed of solids
kept fluidized by the combustion air and the resulting flue gas.
Combustion
temperature is typically controlled in the range of 1500 to l700°F by locat-
ing steam raising surface in the bed and taking advantage of the excellent
heat transfer characteristics and high heat capacity of the fluidized
solids.
Greater than .90% reduction in 502 emissions can be achieved by using
calcined limestone as the bed material.
At fluidized bed combustion tempera-
tures. the lime reacts with the 502 and 02 inl the flue gas to form Ca504.
contacting the sulfated lime with a reducing gas at about 2000°F in a separate
By
fluidized bed, the lime can be regenerated for reuse in the combustor.
This
reductive regeneration also produces a rich 502 by-product stream which can
be used to produce sulfuric acid or elemental sulfur.
The first phase of the program carried out by Esso Research under
Contract CPA 70-19 involved the development of a detailed design for a con-
tinuous fluidized bed combustion - lime regeneration pilot unit - the FBCR
Miniplant.
This unit was designed for operation at up to 10 atmospheres
pressure based on studies by Westinghouse Research Laboratories showing a much
greater incentive for a pressurized fluidized bed combustor used in conjunction
with a combined gas-steam power generating cycle, than for an atmospheric
system.
The FBCR-Miniplant design incorporates a refractory lined combustor
and regenerator with continuous transfer of sulfated and regenerated lime
between these reactors.
The design parameters are:
-------
- 2 -
Combustor
Regenerator
Unit Dimensions
8
Diameter, internal (in)
Height (ft)
12.5
28
5
19
8
Maximum Operating Conditions
8 Tempe rat ure (OF) 1,700 2,000
8 Pressure (atm) 10 10
8 Superficial Bed Velocity (ft/sec) 10 5
8 Heat Released by Combustion (BTU/hr) 6,300,000
8 Cooling Load (BTU/hr) 3,600,000
Maximum Material Rates
8 Air (SCFM) 1,200 90
8 Coal (li/hr) 480
8 Limestone (li/hr) 68
J Natural Gas (SCFM) 10
8 S02 Output (#/hr) 40
Figures 1 and 2 show respectively the overall:system flow plan and
assembly drawing for the FBCR Miniplant.
The design calls for the combustor
to be constructed from 24 inch pipe refractory lined to 12.5 inches internal
diamete r.
It is designed with 5 flanged sections (each 3 feet long), a bottom
plenum for air intake below the removable distributor plate, and an upper bed
expansion section containing the gas disch~rge outlet to the cyclones.
Com-
I.
I
bustor support would be provided by four legs attached to the first section
above the grid.
The reactors would be guided at 8 foot vertical intervals
to accommodate thermal expansion.
The regenerator design features an 18 inch
shell refactory lined to 5 inches internal diameter.
Main fluidizing air for the combustor and regenerator would be sup-
plied by a stationary compressor with a capacity of 1300 SCFM.
The superficial
bed velocity in the combustor and regenerator would be controlled automatically
by differential pressure transmitters and control valves.
In the combustor,
the air would pass through the distributing grid, up through the fluidized
-------
3
FIGURE 1
FBCR MINIPLANT
FLOW PLAN
--_. ---
WATER
..
"'ATaR
0",
CITY ""'AnA
TO Sc.RVeIEA
~
j
W
2
;:
~
,"
tOOll~ --"'1'£R
OVT
.....
~
~
I
8
"""TEA
'"
"'''TIER
~,
Re.S!.RV01"
WA1'I!R RED PuMP
t06P.to\. cr> 60 "...1.
,'--------,
! i
! i
~ IHT!CTtON AUt ~
L -~~.V~"- ~A~~-.J
#.&A. CoOMN.$SOR
8» ."''''.14. \W ',I.I.G
NEAT..
~
A.. C""'''f'"O.
1..-0 s.c"'~. lIe ".'1.1"",
TO 't"uUER
.JIU.
C.S - C:'I"ClONl: ,e.p.-.A.ATOR.
Co Ii. C.OLLI.CTOA. VIO'!I.I.IIL
p,.. POL!.!!. "1.!.OIA
RV - A,!'''ILL VII'!.SIIL.
100fti ~-!
;~~~'~p~=/
, ,
r '£saO ;-~.~~~j;;i r~:~~~AItfG co. I
i U'o)o;tI.. ...
'::;~:~*~i
-------
- 4 -
n
o
1
~
"
d
"
-~-~
~U
ASSEMBLY
-
--
--~i.:
, 0
, ~
I i
/
~Q
17}~~~~:,~-5~g;;.
"ZO-/3-D i
~
-------
- 5 -
bed of solids, and out through two refractory lined cyclones for solids re-
moval before being cooled in a heat exchanger.
The pressure in the combustor
would be maintained at a desired set point by a butterfly valve in the ex-
haust line positioned by a pressure transmitter and controller.
Air for the
regenerator could be electrically preheated for temperature control before
passing into the reducing gas generator located at the bottom of the reactor.
The reducing gas would pass through a ceramic distributor plate supporting
the regenerating bed and then exit through a refractory lined cyclone and a
heat exchanger before discharging through a butterfly valve.
Pressure on
the regenerator and combustor would be kept the same by a differential pres-
sure control system and controller serving to position the butterfly valve
in the regenerator discharge line.
Heat extraction and temperature control in the fluidized bed com-
bustor would be accomplished by boiling demineralized water in 10 independent
loops located in discrete vertical zones of the reactor.
The water flows to
these loops would be controlled by valves whose positions would be automatically
changed to maintain bed temperature in each of the zones.
The steam generated
in these loops would be condensed and returned to a reservoir.
Coal and makeup limestone to the combustor would be fed continuously
from a system designed for controlled solids feeding under pressure.
Solids
transfer between reactors, and discharge of solids from the system (i.e., from
the regenerator reactor) would be accomplished using a pulsed air transport
technique controlled by pressure differentials across and between these
fluidized beds.
This method of solids transfer has been studied in a plexi-
glas Cold Model Test Unit, the CMTU, built for this program.
This unit con-
sists of two 5.5 inch ID transparent vessels connected by two 1 inch ID
transparent hoses for solids transfer.
The CMTU was designed to operate at
ambient temperatures and pressures up to 60 psig.
In the CMTU, a
-------
- 6 -
control panel permitted varying the gas velocities and pressures in the two
vessels, the pulse air rate and pulsing frequency.
Besides verifying the
solids transfer technique and the method of bed pressure control assumed for
the FBCR Miniplant design, the cold model studies permitted visual observa-
tion and measurement of the bed slugging characteristics that could be ex-
pected in the pilot unit when operating with the type of coarse lime particles
for which it was designed.
The second phase of the program involved a study of the factors
influencing the formation and the control of oxides of nitrogen (NO) in
x
fl uidized bed combust ion.
For these experimental studies, a 3 inch ID
fluidized bed combustor and two smaller electrically heated fixed bed
reactors were used.
A fully instrumented analytical train permitted con-
tinuous measurement of NO, S02' CO and 02 emissions.
N02 emissions were
found to be negligible.
The operating factors that were studied included
excess air level, bed temperature, and bed composition.
NO was found to react under fluidized bed combustion conditions
with both CO and S02; the NO emission levels depended on the concentration
of these components in the combustor.
Unreacted lime and sulfated lime were
found to catalyze the reduction of NO by CO with the unreacted material being
the more active.
Aluminum oxide (Alundum) also catalyzed the NO-CO reaction
but
was less active than CaS04'
Based on very limited data, the NO-CO
reaction appears to be second order in both NO and CO with an activation
energy of only 6 to 8 Kcal/g-mole over the temperature range from 1300 to
1700°F.
In the fluidized bed combustor, increasing the CO concentration by
either decreasing the excess air level or decreasing the combustion temperature
produced a decrease in NO emissions.
Thus, with CaS04 (Drierite) as the bed
-------
- 7 -
,
I
I
material, an oxygen level in the flue gas of 4%, and a superficial fluidizing
velocity of 6 fps., NO emissions were 680 ppm at l800°F and 560 ppm at
l500°F.
At l600°F, corresponding NO emissions were 650 ppm at 1% 02 and
565 ppm at 8% 02 in flue gas.
The data that were obtained on the NO-S02reacti~n indicate that
CaO is necessary for this reaction to proceed.
In the absence of oxygen,
fixed bed reactor data indicated an apparent : reaction order of about
0.5 with respect to NO.
This reaction has a negative temperature dependence,
the apparent rate increasing by a factor of about five by decreasing the
temperature from l600°F to l400°F.
This decrease in rate was found to coin-
cide with an increase in CaS03 thermodynamic stability, indicating that CaS03
is probably a reactant in this system.
A reactor configuration designed to yield low NO emissions by
x
promoting the reduction of NO by CO and then the reaction of NO with S02
and lime was tested by modifying the 3 inch fluidized bed unit to permit
its operation as a two stage combustor.
In the first stage, Qperating at
substoichiometric conditions, the NO-CO reaction would be promoted.
The
addition of secondary air at a higher point in the bed would complete the
combustion and provide S02 for reaction with the remaining NO.
Actual
experiments confirmed these expectations.
NO emissions were reduced from
615 ppm at 110 percent stoichiometric air and no staging to 200 ppm when
43% of the stoichiometric air was introduced through the grid in the first
stage and 67% of the air at a point 6 inches above the grid.
-------
- 8 -
INTRODUCTION
The fluidized bed combustion (FBC) of coal is being studied as
a new boiler technique offering the potential of low 502 emissions by
using limestone as the bed material.
In FBC, the bed is kept fluidized
by the combustion air and by the flue gas resulting from the combustion
of the coal.
High heat fluxes are achieved to the steam generating sur-
face located in the bed permitting reduced tubing requirements in com-
parison to conventional boilers.
The excellent heat transfer achieved
in the bed, coupled with its high heat capacity, serves to maintain the
fluidized bed boiler (FBB) temperature at a uniform level typically
controlled in the range of 1500 to l700°F.
Other potential advantages
of the FBB include reduced
steam tube corrosion and fouling, and the
ability to combust low quality fuels.
Within the FBB, limestone is calcined to lime which reacts with
502 and oxygen in the flue gas to form Ca504 as shown by reaction (1).
CaO + 502 + 1/2 02
).
CaS04
(1)
When used on a once-through basis, high limestone feed rates are required
to the FBB if 502 removal
In order to reduce the solid waste disposal burden created by
of 90% or more is
to be achieved.
these high limestone feed rates, a system was proposed by Esso Research
and Engineering Company in which the Ca504 would be regenerated back to
-------
- 9 -
CaO in a separate fluidized bed reactor by reaction with a reducing
gas at a temperature of about 2000:oF.
CaS04 +
{ ~~}
,..
CaO + so 2 +
{ ~~6 )
(2)
The regenerated CaO would be returned to the FBB where it would again
react with S02 and 02.
Esso's Proposed Fluidized Bed
Combustion-Lime Regeneration System
Flue Gas &
Coal FI y Ash
High S02 Gas
to By-Product
Plant
Fluid Bed
Combustor
" / /.. '
" ~;' ."" . (
.. ./ 1/
: ." .'.
Sulfated
Sorbent
Fluid Bed
I Regenerator
Discarded
S orbe nt
Fresh
S orbe nt
& Coal
Re ge nerated
S orbe nt
Fluidizing
Air
Reduci ng
Gas
-------
- 10 -
In a study completed by Esso for the National Air Pollution
Control Administration (l), the following essential features of the
proposed regenerative-limestone FBB system were demonstrated:
.
Removal of over 90% of the 502 formed by combusting coal
in fluidized beds of lime.
.
Reductive regeneration of the sulfated lime to yield an
off-gas containing 7 to 12 mole % 502'
This is a
sufficiently high concentration to permit its conversion
to H2S04 or elemental sulfur with conventional technology.
Good activity maintenance of the lime cycled back and
.
forth between combustion and regeneration.
The make-up
requirement for fresh limestone in a commercial plant was
estimated to be about 15% of that required for once-
through use of this material.
These experimental results were obtained at atmospheric pres-
sure conditions.
Since completing this study, engineering and cost
analyses carried out by Westinghouse (l) for the Environmental Protection
Agency (EPA) have indicated a much greater commercial potential for a
pressurized FBB system when used in conjunction with a combined gas-steam
turbine power generating plant.
Based on this evaluation, the Office of
Air Programs of EPA requested Esso Research and Engineering Company to
study the design of a continuous fluidized bed combustion-limestone
regeneration pilot unit capable of operating at pressures up to 10
atmospheres.
Part I of this report presents the results of this design
study.
-------
- 11 -
Besides its potential for controlling S02 emissions, the
emissions of nitrogen oxides (NO) from FBC units were expected to be
x
low because the low temperature at which the combustion occurs is
unfavorable for the reaction of atmospheric nitrogen and oxygen.
In
fact, however, NO emissions are much higher than predicted from
x
thermal equilibrium for atmospheric nitrogen fixation.
Most of the NO
x
formed in FBC results from the oxidation of the nitrogen organically
bound in the coal.
NO emissions depend on the extent of this oxidation
x
and the subsequent reactions that the nitrogen oxides undergo in the
fluidized bed.
As part of its program for the Environmental Protection Agency,
Esso Research has investigated the effect of different FBC operating
variables on the conversion of fuel nitrogen to NO.
Subsequent reactions
of NO with S02' CO, CaO and CaS04 have also been studied.
The results
of these studies are presented in Part II of this report.
-------
- 12 -
PART I
DESIGN OF THE HIGH PRESSURE FLUIDIZED BED COMBUSTION
LIME REGENERATION PILOT UNIT - THE FBCR MINIPLANT
This part of the report describes the design of the high pres-
sure fluidized bed combustion-lime regeneration pilot unit (the
FBCR Minip1ant) that Esso Research and Engineering Company has prepared
for the Office of Air Programs of EPA under Contract CPA 70-90.
Previous
studies by Esso (1) on the fluidized bed combustion-lime regeneration
system had been conducted at atmospheric pressure in batch pilot unit equip-
ment (see Part II - Section 1 for description of this equipment).'
The design of the FBCR Minip1ant involved three major steps:
(1)
Selecting design conditions and operating parameters
for the unit.
(2)
Detailed design and engineering of the overall unit
and its various sub-systems.
(3)
Verification and modeling of critical design features.
In the following sections of this report, the FBCR Miniplant design is
discussed in terms of these three steps.
-------
L
- 13 -
1.
DESIGN AND OPERATING PARAMETERS
The basis used for the design of the FBCR Miniplant is
summarized in Table 1-1.
Table 1-1
Design Basis for the FBCR Miniplant
Maximum Operating Conditions
Combustor
Regnerator
Pressure (atm)
Temperature (OF)
Superficial Gas Velocity
(ft/sec)
10
1700
10
10
2000
5
Unit Dimensions
Internal Diameter (in.)
Height (ft)
12
28
5
19
The limiting operating conditions for the combustor were set at 10
atmospheres pressure, 1700°F bed temperature, and a superficial velocity
of 10 fps.
These maximum design conditions were based on the engineering
and economic analyses that had been carried out by the Westinghouse
Research Laboratories (l).
A l2-inch diameter combustor size was selected as a basis for
design because this would provide a system which could be constructed
at reasonable cost and within reasonable time while still providing the
essential data needed for future development of the pressurized FBC
system.
At the design conditions, a maximum coal feed rate of 482
lbs. per hour would be possible when operating the combustor with 15%
excess air.
With a heating value of about 13,000 BTU/1b, this coal
rate would correspond to a heat: release rate of 6.3 x 106 BTU/hr.
-------
- 14 -
The internal diameter and operating velocity of the regenerator
were the next parameters to be specified.
As shown in Figure 1-1, these
are not independently adjustable parameters.
They are related to
each other and to the diameter and operating velocity of the combustor
by a sulfur material balance over the system.
A critical factor in this
balance is the S02 concentration of the regenerator off-gas.
Since no
experimental data on pressurized regeneration were available, it was
necessary to assume a value for the 502 concentration in the design of
the FBCR Miniplant.
An assumed value of 4 mole % 502 in the regenerator
off-gas was used for this purpose.
This is a conservative value based
on thermodynamic calculations that were made and which are summarized in
Appendix 1.
Using the 4 mole % S02 concentration as the basis, a 5 inch
diameter reactor and a 5 fps superficial velocity was selected from the
curves shown in Figure 1-1 as representing the best compromise for the
design of the FBCR Miniplant.
A material balance for the designed FBCR Miniplant operating
at its maximum design coal throughput is shown in Figure 1-2.
-------
- 15 -
FIGURE 1-1
RELATIONSHIP BETWEEN REGENERATOR
DIAMETER AND OPERATING VELOCITY
10
en
~
::t:
U
Z
H
.. 8
~
~ MOLE % S02 CONCENTRATION
H
~ IN REGENERATOR EFFLUENT
H
~
~ 2% S02
o
5 6
~
~
t.:)
~
~ 4% 802
4
2
2
4
8
3
5
6
7
SUPERFICIAL REGENERATOR VELOCITY, FEET PER SECOND
Above curves have been 'developed from the equation:
SO
TR . Cc 2
'TC C;02
2 2
.VCDC = VRDR
which is based on a S02 material balance between the combustor and regenerator.
this equation,
In
T = Absolute temperature
CS02 = S02 concentration (based
V = Superficial velocity
D = Reactor diameter
on S content of coal for the combustor)
9
and the subscripts refer to the regenerator conditions (R) and the combustor conditions (C)
-------
Stone
12
TO Cooling H20
3.65 X 106 BT
hr
482
Coal
6.30 X 106
Limes
FIGURE 1-2
MATERIAL BALANCE FOR THE FBCR MINIPLANT
Fly
ash
S02 S
4 1bs./hr. 4
.8 1bs./hr. 2.5 X 106 BTU/hr.
S
1
COMBUSTOR
........ 684 1bs./hr.
glO atm. REGENERATOR
U c::> 1700°F 10 atm.
~
- 2000°F
15,000 BTU/hr Heat Loss
1bs . /hr . v = 10 fps v = 5 fps
-
BTU/hr. 558 1bs. /hr.
1 t
.
tone Air
ned) 1210 SCFM Air 90 SCFM
(Unca 1c1
69.5 1bs./hr.
Fuel
10 SCFM
Basis:
15% Excess Air
4.5% Sulfur Coal
1:1 CalS Ratio
02
o 1bs./hr.
tone
lb. /hr .
I-'
0'\
Stone Discard
25 1bs./hr.
-------
- 17 -
2.
THE FBCR MINIPLANT DESIGN
Figure 1-3 shows the overall system flow plan for the FBCR
Miniplant and Figure 1-4 the assembly drawing for it.
Main fluidizing air for the combustor and regenerator is supplied
at operating pressures to 125 psig by a stationary compressor with a capacity
of 1300 SCFM.
The air flow rates are measured by orifice flow meters and
regulated by differential pressure transmitters and control valves.
The
superficial bed velocity in the combustor and regenerator can be controlled
automatically and independently in this manner.
In the combustor, the
air passes through the distributing grid, up through the fluidized bed
of solids, and out through two stages of cyclone for solids removal be-
fore it is cooled in a heat exchanger.
The pressure in the combustor is
maintained at a desired set point by a butterfly valve in the exhaust line
positioned by a pressure transmitter and controller.
Air for the regenerator can be electrically preheated for tem-
perature control before passing into the reducing gas generator located
at the bottom of the
reactor.
The reducing gas passes through a ceramic
distributor plate which supports the fluidized bed.
The exit gas from
the regenerator is cooled by a heat exchanger before discharging through
a butterfly valve.
The pressure is the regenerator is maintained about
equal to the pressure in the combustor by a differential pressure
transmitter and controller between the combustor and regenerator serving
to position the butterfly valve in the regenerator discharge line.
-------
elf'( ""-'TER
~
g
z
~
o
COOI..ING> 'W}oTER
R~5e.RVOIR
"
WATER FI::ED PUMP
~o G..PM, fP ,"0 P.S I
1------
I
I
! INJECTtOM AIR J
L ~~~~YSTEM- ~.~OP. ~w~
--I
I
I
18
FI CURE
1-3
FBCR MINIPLANT FLOW PLAN
-. "------'._'-~--------'--'-----~~---
_.~-_.._._-
--,----------~-
--,---,~-
WAT£R
'"
WATI!~
0",
OUT
TO SC R uaeER
~
~
t
8
"""-TEl;,
IN
WATER
,,,"
INATEII. O\JT
DJ"CARP
TO &CRV88tR
..!S.S.!..
~ c.YCI..O~..E- 'SE.PAR"To~
c." - COLLECTOR. v~
~i!I,.'5_e_~.E~D~
av - REFIl.l.. vE~set..
G~
AIR (.OMPaESSOI;,
~s.-!:",,.,@ nSf'.s.1.G
I-IEA"TER
J
",Q. :,o""lPI"£SE,CR
1...xJ5'(:"""~ IZS P.'S.',(.
rVVVV'1
-----.- -- ~-_._~~--,
. J
d
I D"'T~ 1
FL.UIOIl.ED C.OAL COMBUSTORI
REGE.NEAATCR FLOW PLAN
~-~J ~~.;. ~~::.-:;.~~~~~~'~~~;:~'~" c:o :
-~----_.
_~-:~'7F.
;:()'
;__n --, --- U.20.14-D
_._--_._~
-------
19
FIGURE 1-4
FBCR MINIPLANT ASSEMBL\
n
o
~
~
"
o
.
~
"
..
z
..
..
~
..
~~
------
..~
I I
..." --
FLCJIOI~ED C.OAL t.CIMaUa-
~~~N&R.ATOA._-- ,..tt.~"'eLY-
-----..-------.
C$SO RESt.I,tr.H - o.:.-..:-.._~:«) co.
ME.>I'.dCAl 0;>.1': ..1
1..8IDI:""'-J
sr;::E~Er--
'-;:' i~~~J1~J'pYJG. NO
.,~.'! --: . - '''20-/3-D
-_. -- ----- .-.
-------
- 20 -
Heat extraction and temperature control in the fluidized bed
combustor is accomplished by boiling demineralized water in 10 separate
loops located in discrete vertical zones of the reactor.
The water
flow
to these loops is controlled by valves whose positions automatically
change to maintain bed temperature.
The steam generated in these loops
is condensed and returned to a reservoir.
Solids transfer between reactors, and discharge of solids from
the system (i.e., the regenerator reactor) are accomplished using a pulsed
air transport technique controlled by pressure differentials across and
between these fluidized beds.
Coal and makeup limestone to the combustor
are fed continuously from a system designed for controlled solids feeding
under pressure.
2.1
Combustion and Regeneration Reactors
The combustion and regeneration reactors constitute the heart
of the FBCR Miniplant design.
The combustor .(see Figures 1-5, 1-6) consists
of a 24-inch steel shell refractory lined to an actual internal diameter
of 12.5 inches.
The overall height of 28 feet was chosen to provide a
bed outage (i.e., dilute phase above the bed) at least equal to the ex-
panded bed height that would be obtained at the maximum operating condi-
tions.
The reactor is designed in flanged sections, with a bottom plenum
for the combustion air, and an upper 'section for discharging the flue gas
to the cyclones.
The regenerator reactor (see Figures 1-7, 1-8) consists of an
l8-inchshell refractory lined to 5-inches internal diameter.
An overall
reactor height of 19 feet provides for bed expansion and reactor outage.
-------
-,.-
.-I
N
A-A
POl
H
~
~
a
H
en
~
~
a
u
[-£'
PeRT LocATION
-£'
c-c
PORr LOGATIONS
..
14)Zc;
D-D
PORT LOGATIONS
I
A
,-I
..
'"
-I
8
"
',:,
'"
-3Z
25
'-r
S',i '
~~
~
~I
~I
~.
~~
a:?
~
"
~
3
'~a !
h; ~.~~'
15' 1..
~1~ ~I~i
~M -".
;a ~ ~ I
~ .~ ~.
1
A
MA T£RIAL
$HELL. Z4~J(..37S.W""L.L sr~EL. PIPE
5HELL ~LAN~l!S . 24" ISO. STEEL R.P; SUP-ON
5HELL. ENDS 'Z4 "STO WALL WELD CAP~
POR T5 NO, 1.1 ~ HALF" PIPE COUPL INC, I
11/0. 2 r".u~u Z!J~2G,27 $~"''''L.~P/~t..COUPLlNG
NO. 24- 3.' J(. ZIt; "WALL)( ~.'L.4. ,sTEEL Pipe ~hi'"
~..SOO-.5TE€L R,~ .sLIP. ON F"LANt;£
NO, ZS, 2'3,", -'Z.Go"x - ZBO'WAtL It G'L.c;. S-EEL
PIPE '..IIT,", ~'~30o..:j7'£EL ~.C: .5L1D.C'N rt.,
No.28'2"x.IS4 WALt.x(;.~1.~5r8El. PIP'EW/rrl
e".!'oo"5r£EL R.r SLIP'ON rLA-A/Gt
NO. 30'S'X..3ZZ.W,ALL X 81c:; STEEL PIPIt WIT/'(
8--S00-5rE£~ RF, SLIP.ON j:"l.A NCIE.
NO.~~ '14 ~ ),,31$. wALL ,I( G;"LG. S-££L PIPE WITH
14 -"3CO".5 rt~L R.~ S(rP 'ON nANqt I
NO.~4 -le.x,S7S.'WALL STEEL PIPE WITH I
12'.150' :5 rEEL. R,~. $1.10'01'0/ rLANGC
NO,"5. ~.... .ZSo.I\'4L.L S~E(I. PIPC ...."rH
Go .J!oo#.srE£l. R.~ ,sI.IP.OA/ ~L,At...;'£
.sPEc/nCAT/CNS
. MArERIAL /l.ND CONSit;luc.r/oN ro CC}/~L.y 11\:/7,., Jo :....:..1
PQc.s5uR£ '/E:3$EI. COD€.
. TO BE rc!jTCD AND A,S,M,£. COOED rot:( 150 p.:),, A,
WOt;;>J
-------
I" PIPE C.OUPl..lioJG>
J,lO~
". " --- ---- ..
22
FIGURE 1-6
REFRACTORY LINED COMBUSTOR
Z: PIPe. C.OV"'LI>Je;.
NO I..,.I/'JINU
/
~
~
I .~:',,-~/""
«I
<1'<
~I
~
- Z~4=----- 3:0".--- --.-3:0- - --- ----3:0 ---1--- 3:0'
--".:p/pe " NO LIAlJ,x,..
3:0'~- --
~
~ /
.' PIP/!. W/7S'ZS Z" LD
~_STP'.oJ""LI. ""'''~
--- - b-O
~
\
l.,re:""E.Ic"t-tT' 50
.L.
- Z:~'.-------I
i
-'~
~I
-------
....:I
....:I
~
,.... tI)
I
.-! ~
C") 0
N ~ ~
;:J
~ ~
H Z
~ ~
~
~
----.-.
- -'.--'-
T'
-r
A
",'Z,/!]
r"
I ~.
,\
'",
I
I
J-
'"
.~
~
I
L
I
!3
A
T'
=,
A'A
POJ...,'T LOC::A nONS
-'~1 G>
~. ~,0
/ ,,'
~ ./ a
""~ ~-- -,----j:-~" 1
',.j-..: "
t
.~
.J
\II .~
9. " -
8.8
P:;;RT LOCATIONS
U:',/9, 23 ..
~---
18
c-c
POR r LOC.A nONS
54'
\
"
'4,'5
e"
..'"
/0/
yY
0.0
PORT i.OCArIONs
I"
I'
h
~
.,
S S
"
: ~~ i: :.~ ...
*~; 'E ~~
..I .t- ",
Ug I"~:I':I
~:I ~ ~I
~ :: fl
'" ~.
III It 1; ..
~
~~
~
~
;;
~
~
5';'.
'"
;;,
MAr£RIAL
SI-I£I..L -18">.:.375 WALl. .s-Eel. PIP,
SriEL.' rLA..\'GE~-/~'I$O" ;j-CEI.. R..c: ~L.':> ::;/1
.:3/'fELL ENOS- I..-J ':;;-."" ".)':'L >-\l':'.~ ~.-":"'.',
po.~r:;, NO, I r;-. i!v 13 -j"~,A:"~ ,<'IPc c..:;.... "':'/\...
vOl4 - 2 HAir O/."C C':o':Jl I,'. ..
"'/0./5,/'&':"4 ,'.,A(:' 0.-..:.-;, "8"" I
No. q - --tf HALF Pi.')~ CCI: ::;)UI.C;
"'0. IB.!I,::,:.'.'1. Go )( .:':,13:'-WAf.l.'~'T.E£- .
.rIP.: ~'./ITH ~ '2:.:, -::Jr.!
:('..- ~LI' '. -.0/ J:"LA t/..,£.:...
110. 19 ,- If... :';'5- 1",..:". sre~L I
PI''''~ ...ITHI2../5CJ"'~T, I
R.r £LI.""O"" r:..;:'.'..,.~
",'0 20 ,'~ 0: .!!7:..; ",/~LL S-.:£I.
1""/::£ WITH 10",/50-:;"'\
R.r. 5L1P'Oll "-LI4N~t: I
NOTE ~
pOllr.:.. 1 rHO!l) I.! 'D~A.'Gi D~/LL 7H,"!u~;H£ll
ALL OTHERS TO 8E AS SHOo'..N
~PECIFIC"'TlON£
M~TER/"L a CON:..Tl:U...ru:;/IJ m 'O""PlV WITH ~ -;Ivf£
P'i'ESO;;t..~£. VE5Stl. CO,jE.
TOB~ TE:JTl!D & A.S,Iof.£. r::O:>EO ~.3R 150 r>.~,.I..J.
WOR."'INrj P~i55UR£. -"T .qoo Dr
1
J
-------
l
,.,.~ PIP&. \I)
5'-",
24 -
FIGURE 1-8
REFRACTORY LINED REGENERATOR
f----
-- (,'-0
-LQ-P1Pt;.....v?:;-'l.e
~. LD.
a z~ "~'-
/,
---.---
, .
!4" PIPE GOvll'~/""G-.-r'fPICAL
~&~,!.EP_'7.::~R.Il....JI'{ E.~(I.
~~T.R~e~o~_!:,_"'-~-."'''u .CA.L
.-- . ---.
/
I
~
--------
/ "
.' ./ " /'
~~H:r '-9-
- - .- ----.-. --- --.
/
- S'-B-
~. .-.-_..- -
1'-8--
'" "t~f_~_PI..IA1<;'-
...!-lP.1...1o..t!.....c.
\
',--
_h ElP~ wj ""~8
12.~ -eWE ",jl./T£CA$T 7S'. 28 5"\.0-
/
z."I....Q..
~IO""EI~:ti
iF--'~.f~ACToR.Y LlN!;JG FOR.u ~1
: Re.&ElJE..QA1':)R ---
j -..
I c.. -, '."L~~" ....-;(~",
! ," ~. }f!. ~. I M
1.',;" */71
We '-- - - l~2J7 ",j I
-------
- 25 -
Bed Support and Gas Distribution Grids
2.2
Figure 1-9 provides the details of the combustor grid design.
This grid consists of 3/8 inch stainless steel plate containing 137 fluidizing
nozzles on a 15/16 inch square pitch.
Each of the 5/8 inch diameter
fluidizing nozzles contains 8 horizontal equally-spaced 5/64 inch holes.
The combustor grid has been designed to provide a pressure drop of about
19 inches of H20.
The regenerator grid is a high alumina porous ceramic plate
that will be sandwiched between the flanges of the main regenerator and
the bottom plenum.
Plate porosity will be picked to give a pressure
drop close to that of the combustor grid.
2.3
Cyclones and Discharge System
In the FBCR Miniplant design, flue gases and entrained solid
particles from the combustor enter a two-stage cyclone separator system
(Figure 1-10).
The solid particles separated in the first stage cyclone
are returned to the combustor near the grid via a dip leg extension pipe.
Solids escaping the primary cyclone enter the more efficient second stage
cyclone where they are separated and discarded by means of a lock hopper
system (Figure 1-13:
two such vessels in series).
This technique permits
the selective removal of fly ash and limestone fines from the system on a
continuous basis.
Gas exiting from the regenerator enters a single stage cyclone
(Figure 1-11) where the entrained particles are collected and discarded by
means of a two-vessel discharge system (Figures 1-12 and 1-13).
All cyclones
are lined with refractory insulation and rated for operation at pressures
to 10 atmospheres.
-------
-----'-'-'---- -.- .-- - --~-- ~_._-- --_.- -.-.--.-----------.---.. ----
--..------ ---
--- -- --- -..- - - - ------- ------.--.-'-' ------. -'-
", '\" '-
",,-" " ,,'
" ',,,- "-. "
"", " '. "-
". <'..... ,,~
.'<.,,-,.
.... ".....'
, """ , .
, '''.
<':-
, '" '
" ....
1---- ---- _.- ---.--------
1'2Y~'
------'-~f
--ASS,r:;,~,-
Sc..At.;:,
. -------~----
(iRF Fc..O L I 'T!:...f:A..~ L 1L:L8-
c.A~r P.E~~T0B:L
I .;..
_.'0/8")(,. 1'3 ~~ .o.'.P:.. DI-.s.T.R1eUTOR,. PLATE_-
~."-,,: I_~'_-
lLz,9.1J:lt> ,
.!:
~
---.i.
I
--1'
~
I~%"
--- ------
/"'---- . -------~
// ~ -----,'-...
/ /./ -~,~}.~ (J.{b~_2>~ ~
" /9 00 0 t() '0 (0 'O--JI::1''--,
. ',(_/:~.,'. :~,' ~~; ~)~, ~, ;? ;; ::. ,~, "-'~'-
I ,.:.J' <:.~ ~. \:;1' "::r#.:. .....-.....:..1 ~/ .:..-4 ,,:.I' ,
i LcJ:-.L:)~60 -{) -0 'O-1(j'- (0).0 .b-9'
/ ,~" '. ", "" "" '.:.. .,..'/ - ,,'I .....{/ I
f-~-4'2>....£'3 -~\ -t'-> -L ,. .4' -i':J.. t~ . ~lf:J ,[,21-- -1..,
',9 ";'" --: \::,....-1- ...~.::::4" -.:i. .~~
~ .... - fA - - ,.t" :." ~i\ ~- "':0. I. -'"!- r~. I '2 ...
\-""-~-"~'V,1:.) Q ~[.:;;;r '9'!oJ 'VI 1..'.1 -q T
unv), ~) ~t) ')--~),u) ",- ,.'" 'f, 'DO '0 t7J>-j
\' "': ' , " " --:~ y ,- '1 ' '. >l'-s;
H"',) '11 ,,) '0 <2 ,,) 0 ~,.. h) .L)-./(':-" (~i.
\ ,../ " '. - -,'" , ,""'- ", . : ',~
\' ,). ,1,.,) <)-~" I"~ '., '21 {+~ t-'. ,:t,
\ ~--~.' ;, ,'~~, ./".,;,-\\ -"":,',, -~~\1':' ',,;)';.',-.~.:~-/-
,,-y,c w t<') Q' -0 ".-" ,- -' '""" "
,! t
~ \ "'''-9 ,) .:1 .:;;\ '~::.,~ ":-,'.' ",jV' '::' '.)
.v ',,-. ~ .'
.: - - 8' ,0/'",,' VI!>, HO,\.ES ", ~)'...L2t-:7--' ,
, "OU"LL'(. SPA<:;.,.,p' , --...j '5'i+-. ----/'//
.--- .__._~,!P~------_.
1")1 ~T~IR/JTOR.. ~LA",=E. - -
~_~~_M.I"
-~.!"-. ~,!3_H_0T .. C.ORRO~IJJ~-.RE..SIST~L..~..LJ..9_Y
AIR DI::"T RiE,I.:TOR, Pl.4..T!:-
~,,:,r.( (.'t: Ilt..WI' E..b C:O-"L C.C.'-A,600,S"TOR.
A~ ::'H('\I.\J'~
i<.\)J BQj;.! \;
I
i
./.of, I
"'O'J .
I.") - -"..J '...
!
~~T8J..~~r-J~. t:'.~_7..L~..__~')~' ~.~
-,:,,:JLL :=-CA..L_~_. I~' RE.~'9....
L
(")
o
@
c:::
en
1-:3
o
~
>
H
~
t::1
H
en
1-:3
:-c
H
b:i
c:::
1-:3
o
~
'"
t""'
>
1-:3
t%j
"%j
H
o
c::: tV
~ (j\
I-'
,
\0
-------
r---- ----
--,
«
!
o ~"""t- (j) :~
'-) ',' ;--> .j, ~- , '", ,1
" , . ,~~«-., \
( '. /'. ""\ ". 8
: f., 'eft, ~;,? . 'l-
\ ~::\~L~'.~, ';iii~:~ :-'" 8/)
.. 'I :, I "'i'. ~ . " '"
(() ,I '\ ,.'..:J ,
" ',' \ <0
1 1>: ,-. --I \~' . " '1'1 ~ -
I '-Jj J) \' ~ '
I ',' , I
I :' \ \ I
1-- -
1 I
L,'-
D/~,
«
-.J
~2~_B'B...
27
FIGURE 1-10
COMBUSTOR CYCLONE
;-
'0" 10 X 11o " on .,,~"C" '. "\
10")<, .3"5 NA~L ~ S PIP&.
I
~----'~-'--ll;--=--+o
J-- - lo~f
--~....._-
Llrr:,'"ualq.I:IT' '5~
PURN:TOft.r
~.!::TL,
.:~i.ll'_.!2bJ....E..!:~GJ!
/s}t"- .,
I~ nHcL 'HGj.f '5TJl.EJJG~
,8RICl
-------
- 28 -
FIGURE 1-11
REGENERATOR CYCLONE
1--
,
,
!
!
I
OJ
---,
«
---- --.
t4";<' c.," 15011.. ~;n
~
, 1(,\
'\JJJ
, , I
,
I
I I
\ I
I
' I I
I ., /
i ' I
" I
./
I /
I I
,.
', ', I '/
,
I I
I I
I I
11 I I I
t. I: : ! .'
1-'- - - - - - + - - r- -, - - -
, I ii'
"
....
,
~lb._$I.~ R£
5!-tP q~U:.L.~N~
'--
l-~~~
i
<{
---1
10'
. ,
"
---~"
,- ,.~ 1<1( 1-11 :!:.TItE...lUil'
S~~ CAo;T Q..I:R.AC-::>""
i
OJ
L
S~C~ION~B'E?.
.se":TI~~__A -~
-- 2""
1
!
\
L~ t::>' oT3).W""\,1.. o;;..f.P~.~~
'''''IO~£.O\J'''.L.I.. S"l P!.E..E....
10''1( 3CC/". ~l'l E.'"
L."" (',.oJ :-: tj...E.
-------
cy..:.:.o~!~ '=".:...'"'AR')lTC"."<1 ~.:::
;::;.=..;= ~'-= f:;,l. -('";:>
1~ :.--
" f': ". - -:.=-.
- '~2C- f-;)
-------
29
FIGURE 1-12
RECEIVER VESSEL
111
. -- ---.-- ------ --- - .
- _u- ~,-
-----------
-I
Jb~ .L50;~ - ~.F' ~:"JP o.J ~:..ANuf:.
\.rre-c,a.,,.r .,$"-2&
_o::.""~T ~E.FJVSO PESE"'~I-i - ENGINEE~NG CO. !
MECHANICAl DIVIS'ON ;
: UNU£N, N~- -
I sr.:1.LE \4". I" I DATE S1.lE~~
..;" 'R.w fJR'........,r '1jz'Z/u OW'. . ;
',.:;,.- :=' , I'ZO'/I.D
- .-- .------+
-------
..1.tL.x..l..~~...$,f,"~I..IP 00 Ft.~~
I-ITl!CA5T $-te CA~T 1:t"......,C.-rok...
r
-~
L
FIGURE
30 -
1-13
RECEIVER VESSEL #2
I-
'\.. - '. " " " '
""'" """ "... \',,' "','" ".
'- ' -"'" ' '-, - ... '\ "',
, " . , , '" "-, '. . . ". .
"-
, " " "... '
. '", '. " '. ". "
" '. -, "', "\.
-- - -----"--
ZlJ"
,
';<~'i: ,
'" "
. '",
, "
/3'
" " " ',.
"" "'. "-. '
" '. .
, '''.,
-'"
'.
''',
~"
"
".
~tO"'J(.'SOl~.}U' 5LIPc»')FL"~L
_18~Jf........!Q'~))_~ll.. REbt.IC.e~
-----
- ------, - ---.-----,,--
I J¥£~~-~~~B-_- !
r ;:;:~~.,:R_~E'::- :-.~,::;;-. ~-I
w,,: -,.'.'lIeAl D ", III
s.=-:l-i- Y:z' . ~~':~'J~~~ ,;: ~H'''~' - ~
1-,.. ~~""... ";'Z/" .;:.r."- I
I IHo~2-:>
. .'
-------
- 31 -
Discharge gases from the cyclones are cooled in heat exchangers
to reduce the exit gas temperatures.
This minimizes the need for re-
fractory lined pipe leading to the scrubber, and lowers the temperature
rating required for the reactor back pressure control valves.
2.4
Combustor Heat Removal
At the maximum operating conditions for which the FBCR Miniplant
has been designed, a combustor cooling load of approximately 3.6 x 106
BTU/hr is required to maintain a l700°F bed temperature.
With the design
calling for a 15 foot expanded bed height, 240,000 BTU/hr/ft of bed must be
removed.
The design that has been developed for this purpose calls for
control of bed temperature by water circulation through 10 individual
serpentine tube loops located in discrete vertical zones of the
expanded bed.
Each loop occupies 18 inches of bed height and consists
of 3/4 inch boiler tubes on a 2-1/4 inch horizontal pitch (Figure 1-14).
The
coolant enters and exits the combustor through 5 special coolant distribu-
tor plates sandwiched between flanges at 3 foot vertical increments in the
lower portion of the reactor.
This arrangement obviates penetrating the
refractory lined shell of the reactor and provides a means of combustor
disassembly for inspection and maintenance.
The combustor cooling water is pumped from a feedwater storage
tank through the tube loops, where a portion of it is vaporized.
The
liquid-vapor mixture then flows through a surface condenser where it is
condensed and returned to the feedwater tank.
Thus,the steam and saturated
water generated in the combustor cooling tubes is condensed, cooled and
-------
$
32
FIGURE 1-14
COMBUSTOR COOLING COILS
<3
1....--------'7}40'
!: }4."
con. SS:C.,IO~ - I
I,
.1-+
.~~';:
L-+
f-'~'l !~'
I HP~I >~-:-
-'t- j- AI- i "I I
,~ If
!! I f
~el' .
//'/'/\\ --t -, 0~'" ""_d~ :,~:'''~; ,';;:;~;,,; >.
/' \..Y ; \.!J /~'-
/0 '- ;f:',0"
/0" ,,14< (2) \
I h'\ ....~ ' e
i 'f-/ ":(,,..,,;"0'''\ \
. ~\
. ',.\
...... ~ '
", j
I'
-~")II. 0,,' ..sALL. 304
SrAINLE5S TOeiNG>
COIL A':)$f.1"\8LY' 1
24'::LS9~8. ~~G~D s"!g.'=.
.f_~6.£. WJ'ltl.;HJ8.. .
~~"TO
PROvIOfL~m lU\J~D F~
.Yz" ")"" PIPE.I!'£. Tl"P. & Pl.ACES
~2."
9' I
~Oll AS-SEMRLY
.~
--~-
'"
, ,
, ,
:1:
, ,
- ~"-
",
c.~
, ,
, ,
: :
:\'i :'-
~,'):,' ~
I
,
I--i
I '
, '
, !
r ~ 'I
I ,
I I
, I
I I
, I
~. -'~
~ ~-i:J~
",~ .. .
, ,
, ,
, ,
I ,
f~
, I
I I
-r'1-
, I
, ,
I ,
I I
, I
-.-
, ,
,-,
- t~--
, I
~::J
COIL ASSEMBLY. 't
S",,",I!. .. .11...':""8\.""1 nC;E~
~~Qt"ITT&D.
,Jore:
F'L"'...~e..-L" ..!:!Qr ~w.l!
'D) SC""Ui. DO'" TO CW.l.JJGE
~... '3oolo.T'O I~oll.,
_"/~~_-
r ~=-~:':I~'~.' ! -;.:~~
FoR 12.}ir;" I.D. REFR..'CTOkY I
LI#JE.O c.O""I::I.)::;.TOK, I
. . ;~,:,(;'I [ .." !
lJfj':- .1. .~ .
. ..'~~.',:~. iJi, " .
~:~~___-'---.:.'~~~'~D I
-------
- 33 -
recirculated to the combustor to maintain a clean, closed cooling water
system.
The fresh make-up water required is demineralized before enter-
ing the feedwater storage tank.
This recirculating arrangement is intended
to minimize cooling tube fouling, thereby maintaining effective heat trans-
fer and extended tube life.
2.5
Coal and Limestone
Preparation and Injection
The design of the coal and limestone preparation and injection
system for the FBCR Miniplant has been provided by Petrocarb Inc. (Figure 1-15).
Petrocarb claims this system to be capable of continuously processing and
feeding the required mixture of coal and limestone to the combustor at a
rate of 550 lbs/hr against a combustor pressure of 10 atmospheres.
The
preparation system dries and crushes the coal to a minus 1/8 inch size
injection quality and delivers the crushed coal to a storage bin.
Coal is taken as received from an existing pile by a front end
loader and placed in the system hopper
from which- it is fed to a rotary
dryer for drying to 1% maximum moisture.
Dryer heat is furnished by a gas
fired hot air furnace, and the moisture laden exhaust gases are vented by an
exhaust blower through a cyclone dust collector and then to the atmosphere.
The collected dust particles become part of the proc~ssed product.
Dried coal discharges from the dryer into the crusher where it
is reduced to minus 1/8 inch size.
It is then conveyed to a 15 ton coal
storage bin.
Limestone can be processed in the same manner in a subse-
quent operation, and stored in an adjacent 2 ton capacity limestone storage
bin.
-------
- 34 -
FIGURE 1-15
COAL/LIMESTONE FEEDING SYSTEM
I - - -- -- -- ------
.'---
..---- --. - ---
------
VE" T
CCo.L
15 Tc\.J
-lIB'
FAf-I
c; 4.S F:Fo..ED
F;jK ,A.;:E
I
I
-~--
INJt."~r:~~
A,~
c'> n:.rKCCA.;;B
---,
I
I
I
I
I
!
I
I
I
1
VEt..IT
I
To PROCI!.SS.,t3SPSISo RE.'2-::I\;'('c... EtJGRG ::S!. UNDEN NJ
REV. DESCRIPTION BY
-------
L--
I
- 3S -
Volumetric feeders deliver coal and limestone in the ratio of
approximately S to 10 parts of coal to 1 part limestone from their re-
spective storage bins to a blender and then to a feed injector vessel.
This mixture of coal and limestone is then transferred pneumatically to
the primary injector vessel upon a demand weight signal from the primary
injector.
After the charge is transferred from the feed injector
to the primary
injector, the feed injector is isolated, vented and refilled in preparation
of a new weight demand signal from the primary injector.
The weight cell
on which the primary injector is mounted is also used to monitor the
materials feed rate to the combustor.
Aerated solids in the primary injector gravity flow through
an orifice into a mixing section where a controlled air stream transports
them into the combustor.
The solids feed rate is controlled by the trans-
port air flow rate and the pressure.
Thus, under emergency conditions,
the coal supply to the combustor can be stopped by cutting off the in-
jection air stream.
2.6
Solids Transfer
Stone is continuously transferred from the combustor to the re-
generator and from the regenerator to the combustor by inducing the solids
to surge into an overflow reservoir immersed in the upper expanded bed of
the reactor (See Figures 1-6 and 1-8).
The solids then flow (by gravity) down
the trans.fer lines into receiving pots located near the grids of the two
reactors.
From these lower receiving pots (Figure 1-16), solids are entrained
-------
- 36 -
FIGURE 1-16
PULSED SOLIDS TRANSFER POT
//6/-0---- (~
. -8/1!/~'''\ . .J\.\
, lit. .' \
1ij- ~t u~1.. ,..--~--~ 0 \
_~~m~J'~ j;
'-. .
0~-'O,/
~/------/
-~~T:" R.EF"R-'C.TORv
I
A"
1'-'"
S"x ,,' 30011.. REDUC.UJC. F'LANCi.E
/
e-' 300110.. RF. SLtPOt.J F~NGE
<-rn
~J~"
I ~
, 1
-~8.~.:~ALL ?I~~
~
/
/
\ .
t:F/"<:~t .
8- .{ ,,,.,,,- ~,O<. ~:~~-
---f," ~(V'"I~RF_:'l;"C'N Il.a.NGE.
~
--- 8-:0:. - 3Z'2~"'U'q..L "'JI:LD UP
I c.o.rE I
~L-SE .o6JR !,E~q
,
I ..~
i
':.":.-':A!..:
.-'..t.;
.. Yt.:..: I ~ ,.'.:'
t21U BR~~T'8I4}1 ,
--,-~~~ZO-{OD:
-------
- 37 -
and transported into the reactor by timed and metered air pulses.
The
pulse interval, frequency and air flow rate regulate the rate at which
these solids are transferred.
Excess solids for discard are also removed
from the regenerator by this technique.
2.7
Bed Level Control
Since the pressure drop across a fluidized bed is directly pro-
portional to the weight of solids in that bed, the design incorporates a
differential pressure transmitter circuit to measure and control the amount
of material in the combustor and regenerator reactors, and therefore their
bed levels at the particular fluidizing conditions.
Stone transfer to
control bed levels is achieved by adjustment of the on-cycle operation of
the pulse feeder mixing chambers.
The regenerator mixing chamber is in-
tended to be pulsed continuously, but the chamber returning solids to the
combustor will operate only when the stone inventory in the regenerator
exceeds its set value.
This regenerator level increase is sensed by the
differential pressure cell as an increase in pressure drop across the bed,
and the pulse air flow solenoid valve will open to transfer solids from
regenerator to combustor.
The stone discard flow from the regenerator, and therefore the
total reactor system solids inventory will be controlled similarly.
Solids
will be discarded from the regenerator to a water cooled receiving resevoir
(Figure 1-12) when the pressure drop across the combustor bed and the pressure
drop across the regenerator bed both indicate high levels.
By this technique,
the bed levels in both the combustor and regenerator can be controlled by
automatically regulating the solids transfer and discard rates.
The regenerator receiving reservoir (Figure 1-12) will also serve
to receive the solids from the regenerator cyclone.
From this reservoir, the
cooled solids will be transferred periodically to a second vessel (Figure 1-13)
capable of being depressurized for solids removal.
-------
- 38 -
2.8
Reducing Gas Generator
A reducing gas generator capable of producing 10,000 SCFH of
gas at 150 psig supplies reducing gas to the regenerator.
The unit is
a 24 inch a.D. carbon steel cylinder internally insulated to create an 8 inch
I.D. combustion chamber.
The insulation is cast to form a 5 inch diameter
discharge nozzle and an off-set shoulder for mating with the regenerator
in such a way as to provide a means for installing a distribution plate
and radiation shield.
The gas/air burner gun is AISI type 309 SS and enters via a
combustion-air-cooled nozzle while the pilot enters via a 2 inch flanged
nozzle on the chamber side.
Two observation windows (2 inch flanged and w/
quartz windows) are provided, so oriented as to allow flame viewing, one
for visual and one for infra-red scanner use.
Pilot and main flame
monitoring is by electronic scanner of weather-proof construction with
flame detector amplifier and relay used to actuate a gas solenoid valve.
A
high tension electric spark for pilot ignition is provided by a
100V/IO,000V transformer.
2.9
Supporting Structure
Figure 1-17 shows the supporting structure designed for the FBCR
Miniplant.
Overall, this structure stands 34 feet tall, is 20 feet wide
and 10.5 feet deep and is constructed of 8 inch wide flange steel beams.
Three platform levels at 8, 16 and 24 feet are provided for servicing the
unit and are reached by a stairway attached to the outside of the frame.
A five ton bridge hoist at the top unit is provided for assembling and
disassembling the unit.
-------
- 39 -
FIGURE 1-17
SUPPORTING STRUCTURE
,-..--.
... n -..-
-- . -- -- .
_.- - -~ .--- ._- --- - - - ---.
0"
--
, - --
o
o
I
, '
,
-t-- i
I
I ,
. : I ! ':"-~ --.-'
C1~1 ;--~ r~. 4-..-r~,' -,-, t:l:'~L' ':1''''
o -li~.-t -. ---I{br=. . - (=-. \Lf --. - --- I I l" "- ---:, '
.1J"EL.:cL..--..._L_,;-, '-~-------;---,r~'----' -~T"'_- j_.~,. " '
i' . I; I 1.11 .: '. ',. " I '. \ '
i: '~::tt j--' 'III:!; "Ii'!!! __~_~L_. .'; I!'i~~-:
, '--r--.tL - ----iJJUr---> :-'-~---~j-J}-- - 'L j-~ '-'1JJ'
~./ I ' n I
, N , -/,.-/)6>r,'-)--d : ! d-1l- '
_.-"":~,.l~; ", ~<---- '---- -..; ~m~ --.; }-m~ f \ I
' ~n---~L ---r~'-' ,; -. . rT- ------ ,
sa; :Y-._~~j~_d ]t- ,;~p I I
-.J . . I .j~~, : (:::~~~
I~- If-!
~. .. -.- - -. =--J~----_. - - -4+-
... 8'0' . --8"0'--
r ~
--~_..._-
._--== i
-- !
'8'e '---- .
'''',''
-- -- - -------r
'-I " : :=-,~~~-...~--::=-=--~~-~~'--~ ",=-={F
i ~ ' : I
J . ~ I
t
,
I
, ~
i
Or
"
; -1
~
~
~
1 ~
I ~ ~
! \ !
I " ;
: ! =-_::.:-=--::..~~~~~--~- _~n'_.__=~=~=:":!
,
i
,
"
Q
U
o
p
~
I
I
I
d
F-
1-1
i~ I
II
I ~---~~~~~------~-~-~~-~-~~------
t
, '
)
, .
"
-1- .
,
~
i
!
~
~ j
): ! :
~~~~-~~-----~=='~~=~---~--~~~1 ~ .~
-10'0;:"
- ---- -- -..
-- --._----
DATE 8£YI!IIONS I
F"I..V'I;)Iz,ED C.OAJ.. c.o~8\J.$roq,/. - i
Pe.c;.ENe.~TOR. .
~":'"!:E..IoJO~" ----.- .
ESse A'E"''''~~ oq r.'!"'I""""'''~ ,..,,\
I Joj..Ct:~'.1CJ... t. .
W:::'I!I. J
I-~E--r;~'-.I~- ..-'
'" J' /J~~~II''''T K/~. -
~ ,i:'ZO./$-D
, --'--
-------
---
~--
-ll'-
~--
I
" 'I
"
:j
o
"-
»"
",
~
"
:j
"
~
OJ'
Ii.
l
I
I
I
~/'
-5'0--
---
/'
I I
I ,
i !
\
5'0,- ..
I"
- -,
-1_,'
.-----;
; 'I
i ' I
I' I
- 1 : i
Iu
II : ':2-
Ii J I
LLi
~~""
I~
- 40 -
FIGURE 1-18
SUPPORT LEGS
VESSEL
1----
"
I'
I
Iii ~
'iL;
~:
-__L
r--
l'
-..L
-'c.
i-~.-;;-~ 1;Z":!_rr:;. ,-2/:)-71
. r8/ii.'oA,OBioJT /6Z0-16-D
-------
- 41 -
The combustor and regenerator reactors are each supported on four
legs connected to the reactors by reinforcing gussets above the grids (see
Figure 1-18).
The steel structure contains guides at 8 foot intervals to
accommodate thermal expansion of the reactor in the vertical direction.
2.10
Miscellaneous Equipment
.
An auxilIary compressor with a capacity of approximately
200 SCFH at 175 psig is intended to supply air necessary
for the pulse feeders, purge lines, combustor start-up
burner and internal fluidizing lines.
The higher pressure
is necessary to assure flow of this special purpose air
into the reactors.
.
A gas burner exhausting into the fluidized bed combustor
is provided for unit start-up at atmospheric pressure.
Fuel for this burner and for the reducing gas generator
is supplied through metering controls from a 300 gallon LNG
storage tank.
.
The system has been designed to allow effluent gasses from
the combustor and regenerator to be combined and passed
through a particulate filter and/or scrubber for fly ash, lime
and S02 removal before discharge to the atmosphere.
.
Pressure, temperature and gas sampling taps are provided at
many locations in the combustor.
-------
- 42 -
3.
DESIGN VERIFICATION WITH THE COLD MODEL TEST UNIT
For the purpose of verifying the design of the pulsed air solids
transfer system used in the design of the FBCR Miniplant, and to permit
visual observation of its fluidization characteristics, a Cold Model Test
Unit (CMTU) was built.
The CMTU is a two vessel fluidized solids system
constructed from Plexig1ass, a transparent acrylic plastic.
The CMTU was
designed to operate at ambient temperature and pressures up to
60 psig.
3.1
Features of the CMTU
Figure 1-19 is a picture of the CMTU.
The unit consists of two
5.5 inch I.D. vessels simulating the combustor and regenerator of the
FBCR Miniplant.
The simulated combustor (i.e., the vessel on the left in
Figure 1-19) is 18 feet tall and was assembled from 3 flanged sections.
The regenerator is made of 2 flanged sections and stands 12 feet tall.
Both vessels have bottom plenum chambers and grids for distributing the
fluidizing air and single stage aluminum cyclones for removing entrained
solids.
These solids can be returned to the fluidized beds or collected
as desired.
One inch I.D. nylon reinforced transparent PVC hose is used for
solids transfer between the two vessels.
This permits visual observation
of solids movement in these lines which connect the upper part of the
vessel transferring the solids to the lower part of the vessel receiving
them.
The small receiving-injection pots to which the bottom of these
lines are attached are equipped with fluidizing and pulsed air.
In some
of the solids transfer studies, the use of an overflow collector was
-------
- 43 -
FIGURE 1-19
THE COLD MODEL TEST UNIT
-i..:aJ1I8Ii
- .
i
'" ~,
1.-
..- - ~
~ ...
.
---
...-~.'.
.
-------
- 44 ...,
investigated as a means of creating an enlarged solids reservoir in the
transmitting vessel.
In the FBCR Miniplant, such a reservoir would serve
to insure a continued solids seal in the transfer legs and minimize back-
flow of gases between reactors.
The enlarged reservoir, shown on the
regenerator vessel in Figure 1-20, was constructed of a 6 inch I.D. aluminum
tee with a plate on the inside to serve as a solids baf~le and a plexiglass
window on the outside for observation of the solids.
In addition to studying solids transfer techniques, the CMTU
was used to determine what effect the hair pin heat transfer loops
designed for the FBCR Miniplant would have on the fluidization character-
istics of the combustor.
Figure 1-21 shows a simulated heat transfer.
loop sectio~ that was used in the CMTU.
It was fabricated of 3/8 inch
aluminum as a prototype half-scale version with the same configuration'
and pitch as the FBCR Miniplant design.
Gas flow rates to each of the two reactors of the CMTU could
be varied.
A control panel situated to the left of the plexiglass vessels
permitted measurement of their flow rates and of their individual bed
pressure drops.
It also controlled and measured the pressure differential
between beds, and provided a means of varying the on-off pulsing times
and the air flow rates for the solids transfer pots.
3.2
Solids Transfer Studies
The pulsed gas-solids transfer technique that was selected
for the design of the ~BCR Miniplant has bep!l. uc:po sllcr:.essful!y at the
Esso Research Center in England (1) for transferring solids between the
two atmospheric pressure chambers of its continuous CAFB pilot unit.
The
-------
- 45 -
use of this technique in pressurized systems had not been demonstrated
however, prior to the CMTU studies.
Based on the results obtained in
the CMTU studies, it is now clear that:
1.
Solids can be transferred between fluidized bed reactors
under pressure at a controlled and adjustable rate using
the pulsed-air technique.
This has been verified at pressures
up to 40 psig, and fluidizing velocities up to 10 fps.
2.
The solids transfer rate
can be controlled by changing the
on-off pulse cycle time or the pulse air flow rate.
3.
The bed,leve1 control system designed for the FBCR Mini-
plant works.
This technique, which consists of continu-
ous1y pulsing one of the transfer receiving pots and
controlling when the other is pulsed by the pressure
differential between beds,
is
able to maintain constant
bed levels in both vessels of the CMTU without any
adjustment to the controls once they were set.
4.
The use of a solids reservoir decreases the likelihood
of inadvertently emptying the transfer legs completely of
solids.
5.
An increase in system pressure decreases the solids
transfer rate that can be achieved with a given pulse cycle
time and gas flow rate.
6.
Solids ranging in their properties from glass beads (7
to 10 mesh and 110 1bs/ft3 bulk density) to limestone
(-6 mesh and 85 1bs/ft3 bulk density) can be transferred
using the pulsed gas technique.
-------
- 46 -
FIGURE 1-20
SOLIDS TRANSFER RESERVOIR
I
I"
I
1
.,.,
...tt
..
\
II ~n'. i
~__f
I
-I
-... -
-------
~~ ," ~..
~.. .- .-... ,......" " .
- 47 -
FIGURE 1-21
SIMULATED HEAT TRANSFER LOOP
-------
- 48 -
3.3
Solids Fluidization and Entrainment
3.3.1
Predicted Regenerator Bed Slugging Height
The fluidization and entrainment characteristics to be expected
in the regenerator of the FBCR Miniplant were studied in the CMTU.
The
taller of. the two CMTU vessels was used in these studies but without the
simulated heat transfer section.
The range of parameters investigated
is shown in Table 1-2.
Table 1-2
Parameters in CMTU Entrainment
and Fluidization Studies
Superficial Gas Velocity,
Settled Bed Height, ft
Vessel Pressure, psia
. Fluidized solids
ft/sec
4, 6, 8 & 10
2 & 3
15, 40 & 60
7 to 10 mesh glass
-6 mesh limestone
beads &
In the way of qualitative observations, slugging of the
bed was found to be worse with the spherical glass beads than with the
wide cut limestone.
The quality of the fluidization also deteriorated
with increasing settled bed depth, more so for the limestone than for the
glass beads.
At a settled bed height of 3 feet, the slugging behavior of
the limestone and glass beads became similar.
When the maximum slugging
height exceeded the vessel outage and slugs of solid particles entered the
cyclone, bridging of the cyclone solids outlet tube would occur resulting
in a tota110ss of . solids capture capability.
-------
- 49 -
Figures 1-22 and 1-23 show the mean and maximum bed heights obtained
in the CMTU with the -6 mesh limestone material as a function of gas velocity,
settled bed height, and operating pressure.
The particle size distribution
for the limestone is shown in Figure 1-24.
Using the data shown in Figure 1-22, equation (1-1) for predicting
the average slug height was derived using a least squares regression
technique.
H- = 0.265 hO.S pO.4 (U-v.)0.82
s
(1-1 )
H- = average slug height above the settled bed
s
in fee t
h = settled bed height in feet
p = fluidizing gas pressure in psia
U = superficial gas velocity in ft/sec
v = incipient fluidization velocity in ft/sec
Since the regenerator diameter in the FBCR Miniplant design approximates
that in the CMTU, equation (1-1) can be used directly to approximate the re-
generator height that will be required for the continuous pilot unit.
Con-
sidering the maximum operating conditions for which the FBCR Minip1ant has
been designed (see Table 1-1) and assuming a settled bed height of 3 feet,
equation (1-1) predicts an average slug height of 11 feet for the regenerator.
As shown in Figure 1-23, the maximum slug height will be about 3 to 4 feet
higher than this.
Thus, the FBCR Miniplant design which calls for an over-
all regenerator height of about 16 feet from the grid to the outlet should
-------
- 50 -
satisfy the most extreme slugging conditions expected in the pilot unit.
If this turns out not to be the case, a baffle could be easily installed in
the regenerator prior to the cyclone to break-up the slugs of solids.
3.2.2
Predicted Combustor Bed
Slugging Height
The use of equation (1-1) for determining the slugging height in
the 12.5 inch combustor designed for the FBCR Miniplant should result in
too high a predicted value.
This is so because its larger diameter, and the
baffling provided by its heat transfer loops, should make this reactor much
less conducive to solids slugging than is the regenerator.
At the extreme
operating conditions for which the FBCR Minip1ant has been designed (see
Table 1-1), and a settled bed height of 2 feet, this correlation predicts
a mean slug height of about 17 feet.
Thus, the maximum slug height obtained
in the combustor should be well within the 23 feet provided by the pilot
plant design.
-------
22
20
18
~
~
~
~
16
~
H
c:x:
~ 14
::E:
o
c:x:
~
A 12
~
::I:
~
H
~ 10
~
Z
H
~
~
~
~
tf)
~
~
~
~
~
~ 15 psia
. 40 psia
. 50 psia
---
FIGURE 1-22
MEAN SLUG HEIGHT OBSERVED IN CMTU STUDIES
3.0 FT. BED HEIGHT
2.0 FT. BED HEIGHT
CMTU BED MATERIAL
GROVE-1359
4
2
o
2
- - -- - -=-......~
..... -- ~ ~ --
."",. ."",. ~ ---
.,,-- ."",.
." ."".---
". ."".,r
" ". --- ."",.
~ ."",.
.-
V1
t-'
4
6
VELOCITY, FT./SEC.
8
10
-------
20
18
~ 16
~
....
Q 14
H
p::;
o
::E:
o 12
p::;
....
ft
Eo;
G 10
H
~
;I::
~ 8
H
o
o
:::>
Ci3 6
~
::E:
H 4
~
2
o
2
FIGURE 1-23
MAXIMUM SLUG HEIGHT OBSERVED IN CMTU STUDIES
MAXIMUM HEIGHT BEFORE DISCHARGE
TO CYCLONES WAS 16 FEET
tt
~JJ
~~
~~
~~
~~
~~
~~
4
6
8
VELOCITY, FT./SEC.
V1
N
10
-------
7000
6000
5000
4000
3000
2000
:1
..
J:I:j
~.
~ 1000
~
~
~
U
H
H
~
P>4
FIGURE 1-24
PARTICLE SIZE DISTRIBUTION OF BED
MATERIAL USED IN CMTU STUDIES
..
. 2.0. FT. BED MIX (STARTING MATERIAL)
. AFTER 40 ps ia SERIES
. AFTER 15 psia SERIES
---
AFTER 60 psia SERIES
CMTU BED MATERIAL
GROVE 1359
EACH SERIES OF RUNS
LASTED 1 HOUR
400
300
200
100
10
20
30
o
40
50
CUMULATIVE WEIGHT) PERCENT LESS THAN
60
-------
- 54 -
P ART II
FACTORS AFFECTING NOx FORMATION
AND CONTROL IN FLUIDIZED BED COMBUSTION
Part II of this report deals with the experimental study of
NO
x
formation and control in fluidized bed combustion that was carried
out as part of Contract CPA 70-90 for EPA's Office of Air Programs.
The
experimental equipment and procedures used in this study, data obtained,
and the analyses of the results are presented in the following sections.
-------
- 5~ -
1. EXPERIMENTAL APPARATUS,
MATERIALS AND PROCEDURES
The experimental equipment used for this study of NO
x
formation
and control consisted of a fluidized bed coal combustion unit and two fixed
bed reaction units.
1.1
Fluidized Bed Coal Combustion Unit
A schematic diagram of the Ess6 fluid bed combustion unit is shown
in Figure 2-1.
The three primary components of this unit were:
(1) the
coal feeding system, (2) the fluidized bed reactor, and (3) the gas cleanup
and analysis system.
1.1.1
Coal Feeding Equipment
Coal was stored in a small hopper and gravity-fed to a vibrating
low rate solids Vibro-Feeder.
The output of the feeder was connected by
pneumatic tubing to a 1/4 inch stainless steel tube, extending through the
reactor grid approximately 1/8 inch into the bed.
A small baffle was located
directly above the coal inlet to promote dispersion.
The feed rate of the
coal was controlled by adjusting the vibration intensity of the feeder.
1.1.2
Fluidized Bed Reactor
A schematic diagram of the fluidized bed reactor used as the coal
combustor in this study is shown in Figure 2-2.
This reactor consisted of an
air preheat section, a reactor section, and a disengaging section.
The
entire reactor was constructed of an Incoloy 800
a 11 oy .
The air preheat section was a three inch diameter, eight inch
high section, packed with 1/4 inch ceramic cylinders.
It was attached
directly to the reactor section.
Beaded electrical heaters supplied the
necessary heat to preheat the air to 600°F.
The air preheat temperature
was automatically controlled at the desired level.
-------
Vent
WTM
Instrument Calibration By-Pass
Figure 2-1
ESSO ~LUIDIZED BED
COMBUSTION UNIT
Feeder
Scale
Coal
Hopper
and Feeder
Gas Rotameters
N2 CO S02
02 NO Air
Cyclone and Filter
Air
Condenser
Reactor
Water
Condenser
Condensate
Refrigerator
IR S02
Anal yzer
IR CO
Analyzer
NOx Analyzer
Polarographic
02 Analyzer
Intermittent
Gas Sampler
WTM
v-
a-
WTM
-------
I
18"
7" p.,...Al'
DP
48"
Cooling
Coil ~
TI
Electric I
Resistance ------
Heateue
, I
Fluidizing --.
Air 1/411 T
- 57 -
f Flue
Gas
.-3/4" P
Limestone
~Addition
Port
--3/4" P
Coal
Inlet
t
Figure 2-2
FLUIDIZED BED COMBUSTOR
Solids
Removal
Screw
-------
- 58 -
The reactor was a three inch diameter, four feet high Incoloy
tube.
Air distribution was accomplished by six bubble caps, with eight
5/64 inch outlet holes in each bubble cap.
The bed was preheated by use
of
beaded, resistance wire heaters.
Bed temperature was controlled by
use of an external cooling water coil.
The water flow rate to the cooling
coil was manually adjusted to yield the desired temperature.
The bed was
insulated to limit heat loss to the atmosphere.
Average bed temperature
was determined by five thermocouples in the bed.
Pressure drop across
the bed was continuously determined by use of a water manometer.
With-
drawal of bed solids was accomplished by use of a screw conveyor and fresh
limestone was batch charged to the unit by use of the inlet port located
three feet above the distributor plate.
The disengaging section was a seven inch diameter, 18 inch high
section,
in which the gas velocity was reduced to allow some entrained
particles to be returned to the bed.
During the course of the study on nitrogen oxides, the fluidized
bed reactor was modified to permit staged air addition to the combustor.
This modification, shown in Figure 2-3, consisted of adding secondary air
ports at 0.5 and 2.5 feet above the primary air distributor grid.
The
purpose of this modification was to study the effect of staged combustion
on NO
x
emissions.
1.1.3
Gas Cleanup and Analysis System
The gas cleanup and analysis system that was used is also shown
in Figure 2-1.
The gas exited the reactor through a 3/4 inch uninsu1ated
pipe.
The temperature of the gas dropped to approximately 300°F before
entering a glass cyclone and a fritted stainless steel filter.
The gas
leaving the filter was essentially free of entrained particulates.
-------
- 59 -
FIGURE 2-)
STAGED CQMBUSTION REACTOR
Secondary Air
lover Bed
/ r
I
1/4" T 2'6"
I
\ i
I
I
i
I
AL
I
- 6"
~
Secondary Air In Bed
-------
- 60 -
The cyclone and filter were enclosed in an electrical oven to prevent any
moisture condensation.
The particulate-free flue gas was then passed through an air-
cooled condenser and a water-cooled condenser to remove most of the water
vapor.
Approximately 50 percent of the flue gas was then metered by a
large wet test meter and vented; the remaining flue gas ( 0.5 CFM) was
diverted through a refrigerator which lowered the dew point of the gas
to 35°F before it was sent to the gas analysis equipment.
The gas analysis
equipment included:
Beckman Model NDIR3l5 Analyzers for S02 and CO, a
Beckman Model 715 polaragraphic analyzer for 02' and a Whittaker polara-
graphic NO analyzer.
x
During the program, a Dupont Model 461 NO analyzer
x
and a Beckman Model NDIR3l5B NO analyzer were also used.
The Dupont
analyzer measures both NO and N02'
N02 is measured directly by absorption
of visible light while the NO is measured by oxidizing the sample gas to
N02 under 60 psia oxygen pressure and measuring the resulting N02 concen-
tration.
A single determination of both NO and N02 typically required
five minutes with the Dupont analyzer.
1.2
Fixed Bed Reactors
1.2.1
2.5 Inch Reactor System
A schematic diagram of the basic fixed bed reactor system used
in
this study is shown in Figure 2-4.
This system was equipped to operate
with either upward or downward flow of gas.
Cylinder gases of NO, S02 and
CO were used to investigate possible reaction systems occurring in the
Esso FBC.
Inlet and outlet NO, S02' 02 and CO concentrations were mea-
sured using the instrumentation described in Section 1.1.3.
Most of the studies
made with this system were carried out in a 2.5 inch O.D. (2.25 inch I.D.)
ceramic tube externally heated by an electric furnace.
-------
VENT
BECKMAN
°2
ANALYZER
BECKMAN
S02
VENT ANALYZER
NO-N02
VENT ANALYZER
- 61 -
Figure 2-4
FIXED BED REACTOR
(Shown for Upflow Operation)
ELECTRICAL
HEA TER
GAS
MIXER
B
VENT
\
"- ALUMINA
CYLINDERS
FOR BED SUPPORT
AND PREHEA T
I BPR I BACK PRESSURE
REGULATOR
g ROTAMETER
NO 1-0 PRESSURE GAUGE
S02 IN N2 ~ BALL VALVE
N2
AIR ~ NEEDLE VALVE
-------
- 62 -
1.2.2
1 Inch Reactor System
In additfon to the 2.5 inch ceramic reactor, a 1 inch O.D.
stainless steel reactor (3/4 inch I.D.) was used to carry out fixed bed
reaction experiments on an even smaller scale.
The same gas blending and
analysis equipment was used with this reactor as was used with the 2.5 inch
reactor.
1.3
Feed Materials
1. 3.1
Limestone
The limestone that was used in this study was obtained from the
Grove Lime Company, Stephen City, Virginia.
The chemical analysis of
this stone is shown in Table 2-1, and the various particle size distribu-
tions used are given in Figure 2-5.
Calcined stone was used primarily,
but not exclusively, in this study.
The normal calcining procedure and
conditions are given in Appendix 2.
The bulk and particle densities of the stone are given in
Appendix 3A, along with the experimentally determined minimum fluidizing
velocity
curves.
Table 2-1
Composition of Limestone
Used in Esso FBC Program
Limestone - Grove
N-1359
Component Wt Percent
CaO 97
MgO 1.16
Si02 1.07
A1Z0) 0.29
-------
5000
4000
3000
2000
~ 1000
M" 900
~ 800
Q)
E 700
.~ 600
o 500
..
Q)
(j 400
.,...
...
M
~ 300
~
200
100
90
80
70
60
2
.. 63 ..
Figure 2-5
PARTICLE SIZE DISTRffiUTIONS OF LIMESTONE
AND COAL FEEDS TO ESSO FBC
Lime N-1359
o
o
Litne N-1359-o
1)- 0 (Drierite)
caS 4
~
u
coal
-0--
-
---0-
98
5
10
15 20
40
30
Cummulative Weight Percent Less Than
-------
- 64 -
1.3.2
Coal
The coal used was a high-volatile (A) bituminous coal from Northern
West Virginia.
It was obtained from the Humphrey Preparation Plant,
Christopher Coal Company, a division of Consolidation Coal Company, Inc.
The Bureau of Mines at Morgantown, West Virginia, obtained the coal from
the pr~paration plant and subsequently shipped it to Esso.
The chemical
analysis of this coal is given in Table 2-2.
The particle size analysis of
the coal is given in Figure 2-5.
Table 2-2
Composition of Coal
Used in Esso FBC Program
Coal - Pittsburgh Bituminous (A)
. .' Proximate Analysis
Component Wt. Percent
Ultimate Analysis
Component Wt. Percent
Moisture
Volatile Matter
Fixed Carbon
Ash
1.2
6.5
53.3
9.0
Ash
H
C
N
S
° (by diff.)
9.0
5.3
74.1
1.4
3.0
7.3
1.3.3 . Alundum and CaS04
Abrasive-grade Alundum (20 mesh) was used as a bed material in
selected runs.
(The minimum fluidizing velocity and the bulk density of
this material are given in Appendix 3B.)
CaS04 (Drierite) was also used as a bed material.
fluidizing velocity and bulk density of the caS04 also are given in
(The minimum
Appendix 3B.)
The particle size distribution of CaS04' was the same as
the particle size distribution of the smaller lime, as shown in Figure 2-5.
-------
- 65 -
1.4
Experimental Procedures
1.4.1
Fluidized Bed Combustor
,.... ,
The startup of the Esso FBC was accomplished by batch addition
of the desired amount of bed material to the unit and preheating the bed
to approximately l200°F.
Upon reaching this temperature, coal feed to the
unit was initiated.
When the desired operating temperature of the unit
was approached, the coal feed rate was adjusted to the desired operating
rate and cooling water was introduced to the cooling coil for control of the
bed temperature.
A steady combustion condition was assumed to be achieved
when both the bed temperature and 02 level in the flue gas stabilized.
Samples of the cyclone solids and bed solids were taken intermit-
tently as the run progressed.
At the termination of a run, samples of the
bed and cyclone solids were also obtained.
These solid samples were analyzed
for carbon, sulfate, and sulfur, as desired.
1.4.2
Fixed Bed Reactors
The experimental procedure followed in using either one of the
fixed bed reactors was as follows:
1.
The reactor would be charged with the desired bed material
and electrically heated to the desired temperature.
Nitrogen and argon would be used as purge gases during
"
the heatup period.
2.
The inlet composition of the reactant gases (blended
mixtures of NO, S02 and CO) would be determined.
The reactant gases would then be fed to the fixed bed
3.
reactor (either upflow or downflow), and the outlet gas
compositions continuously monitored.
-------
- 66 -
2. GENERAL CONSIDERATIONS OF
NO FORMATION AND CONTROL
x
IN FLUIDIZED BED COAL COMBUSTION
The oxides of nitrogen, NO and N02 (collectively referred to as
NO ), can be formed in a combustion process by the direct combustion of
x
atmospheric nitrogen and oxygen as given by equation (2-1).
As shown
N2 + 02
)-'
2NO
(2 -1)
,
in Figure 2-6, the equilibrium NO concentration in this system is low, less
than 100 ppm, at commonly employed temperatures in fluidized bed combustion
of 1500 to 1700°F.
Hence, it was originally anticipated that NO emissions
from fluidized bed combustion systems would be low.
Experimental studies at Esso have shown, however,
that NO
emissions greater than 1000 ppm can be obtained in a fluidized bed combustor
operating at l600°F.
Such concentrations are considerably above the
equilibrium concentration for NO shown in Figure 2-6 at a temperature of
1600°F.
Research conducted at Argonne National Laboratories (~) has shown
that the oxidation of fuel nitrogen is the primary mechanism for NO formation
in fluidized bed combustion.
This oxidation can be conceptually represented
by equation (2-2).
N
fuel
+ 1/2 02
~
NO
(2-2 )
The rate of reaction (2-2) is very fast compared to the rate for the
reverse reaction (2-1) and it is conceivable that all of the fuel nitrogen
is initially converted to NO.
The concentration of NO appearing in the
flue gas would then be dependent upon subsequent reaction of the
-------
250
200
K 150
Co
..
u
z
o
u
o
z 100
50
o
1000
1340
Figure 2-6
THERMODYNAMIC EQUILIBRIUM NO CONCENTRATION
Methane
Combustion
o = 410
2
o = 210
2
1100
1200
TEMPERA TURE, oK
1700
TEM PE RA TU RE, of
1300
1520
1880
0'\
~
-------
- 68 -
NO in the bed.
Prior to the studies described in this report, the following
observations had been made by various investigators concerning NO emissions
from fluidized bed combustors:
1.
The NO emissions from the FBC with a partially-sulfzted
lime bed are 20 to 40 percent lower than the emissions
(under similar operating conditions) using an inert bed. (~)
2.
The addition of C0304 to a
FBC, operating with an inert bed
material, causes an increase in NO emissions, indicating
that C0304 may have an accelerating effect on the oxida-
tion of fuel nitrogen (~).
3.
The NO emissions from a FBC with a lime bed (zero.
sulfation) are higher than corresponding emissions with
an inert bed.
These higher emissions may reflect a catalytic
effect of the lime on the rate of oxidation of fuel
nitrogen (...1).
4.
NO and S02 react over a partially sulfated lime bed to
yield lower NO emissions.
No reaction occurs over an inert
bed or CaS04 (...1).
These results served as the starting point for the basic and
applied studies described in Section 3 of this part of the report.
-------
- 69 -
3.
EXPERIMENTAL RESULTS
3.1
Effect of Bed Temperature and Excess Air on NOx Emissions
NO emissions resulting from the reaction of atmospheric nitrogen
x
and oxygen in conventional combustion equipment are known to be very much
influenced by the oxygen concentration of the flue gas (i.e., the excess
combustion air level) and the temperature in the combustion zone.
On the
other hand, in fluidized bed combustion where most of the NO
x
is formed by
oxidation of the organic nitrogen in the coal, detailed studies have not
I
been made to define the effect of excess air level and bed temperature
on the conversion of the fuel nitrogen to NO .
x
The initial series of FBC
experiments were designed therefore to obtain such data.
Alundum
(Alumina) and Drierite (CaS04) were used as the fluidized bed materials
in these experiments since these materials, unlike lime or limestone,
would be inert to reaction with S02 formed from the oxidation of the
sulfur in the coal.
Consequently, the S02 concentration would remain
constant and would not be an influential variable in these studies.
This
was essential since it had been shown in previous work at Esso Research (1)
that S02 and NO reacted in the presence of lime at fluidized bed combustion
conditions.
The use of Alundum as a bed material provided a base condition to which
emissions with lime and CaS04 could be compared.
CaS04 was used since
it is the product formed when lime is used as the fluidized bed medium
for S02 emissions control.
formation of CaS04 would have on NO emissions would be maximized by its use
Therefore, any catalytic effect that the'
as the bed material.
-------
- 70 -
Figure 2-7 shows the NO emissions obtained operating the fluidized
bed combustor with the lime, alundum and CaS04 bed materials.
The N02
emissions were essentially zero in these tests, and for that matter,
in
all subsequent fluidized bed combustion tests that were carried out.
The
NO emissions using CaS04 were constant at 615 ppm.
With the Alundum bed,
the NO emissions were higher at 725 ppm, but still constant.
The S02 emis-
sions in both these test were constant at a level of about 2500 ppm, co!respo~ding
to complete conversion of the fuel sulfur to S02'
the lime bed were the highest observed at about 920 ppm.
Initial NO emissions with
As the run progressed,
however, these emissions appeared to asymptote to about the same NO level
obtained using the CaS04 bed.
3.1.1
Bed Temperature
The NO emissions from the Esso FBC were monitored over the range
of bed temperatures from l400°F to 1800° Fwith Alundum and CaS04 as the
bed materials.
Complete data reduction for these runs is given in
Appendix 4.
are shown 1'n Figure 2-8, as function of bed
The NO emissions
temperature, for these two bed materials.
The NO emissions using
the Alundum bed, were consistently higher than the corresponding emissions
with the cas04 bed, indicating that some sort of catalytic. reduction.
(or reaction) of NO was occurring over CaS04.
between the NO emissions obtained with the Alundum and CaS04 beds increased
Interestingly, the difference
as bed temperature was decreased.
This result indicates that whatever
phenomenon was acting to reduce NO emissions over the CaS04 bed was being
accelerated at low bed temperatures.
-------
1000
- 71 -
FIGURE 2-7
NO EMISSIONS USING DIFFERENT BED MATERIALS
800 ALUNDUM
~
p.,
p.. CaSO
600
(J)
z
0
H
(J)
(J)
H 400
~ T = 16000F
o U = 6 fps
z 02 (FLUE GAS) = 4'70
200 Ho = 6 INCHES
o
o
0.5
1.0
1.5
RUN TIME. HOURS
2.0
-------
Figure 2-8
NO EMISSIONS AS A FUNCTION OF BED TEMPERATURE
700
ALUNDUM BED
--...L
.--
--
800
-
~ 600
-c
-
E
Co
Co
.. 500
(/)
z
o
~ 400
~
UJ
~ 300
-....J
N
200
U = 6 fps
Ho = 6 In.
o = 4<}'o
2
= FBC
- - = FBC with 250 ppm NO added
to fI uidi zing gas
100
o
1300
1400
1500 1600
BED TEM PERA TURE, OF
1700
1800
-------
- 73 -
One possible explanation for this apparent acceleration at low
temperatures stems from the observation that the CO emissions increased
in these experiments as bed temperature was decreased.
The CO
emissions
were 250 ppm at l800°F and 5200 ppm at l400°F.
Therefore, it is likely that
the lower NO emissions obtained with CaS04 (as compared to Alundum) could
have been the result of the reduction of NO by CO, with CaS04 acting as
the catalyst. Although a similar surface reaction could also be occurring
over the Alundum, the much lower surface area of the Alundum would not be
expected to cause any significant catalytic enhancement of the reaction.
In order to obtain additional information on the effect of
temperature on the formation of NO in the fluidized bed combustor, experi-
x
ments were performed adding 250 ppm NO to the fluidizing gas of the Esso
FBC for each of the bed temperatures examined.
The concentrations of NO
in the flue gas following this addition are shown in Figure 2-8, by the
dashed lines.
The increase in NO concentration in the flue gas (following
NO addition to the fluidizing gas) was generally small, indicating that
much ot the 250 ppm NO added to the fluidizing gas had reacted as it
passed through the reactor.
In order to compare the fate of the nitrogen in the fuel to the
fate of the nitrogen added as NO, the following calculations were made:
1.
The fractional conversion of fuel N to NO was calculated.
2.
The fractional conversion (retention) of added N (as NO) to NO
was calculated.
Both of these calculations are illustrated in Appendix 5.
-------
- 74 -
As shown in Figures 2-9 and 2-l0~ the fractional conversion of
fuel nitrogen to NO was consistently greater than the fractional conversion
of the added nitrogen to NO (i.e. more decomposition or reduction of the added
NO occurred).
This result is believed to reflect the greater reduction of the
NO added to the fluidizing air as it passed through the highly reducing zone
that existed above the grid of the Esso FBC.
This reducing zone existed because
the coal was fed to the bed at single point in this unit~ i.e.~ through the
center of the grid.
This implies that increasing the number or extent of
localized reducing zones within the bed would cause a decrease in NO
emissions.
3.1.2
Excess Air
The NO emissions from the Esso FBC were determined as a function
of the excess air level used for the combustion over the range of 1 to 8 per-
cent 02 in the flue gas~ using Alundum and CaS04 as bed materials.
The com-
plete results of this set of experiments are given in Appendix 4.
The actual
and normalized (3% 02) emissions are shown as a function of percent 02 in the
flue gas in Figures 2-11 and 2-12 for Alundum and CaS04 beds~ respectively.
-------
...-
....
<1>
-
1:S 0.5
o
~
o
2 ....
o ~ 0.4
(1)0 <1>
a::: ~
2....
~o~
21-~0.3
o ..., .,...
U~~
...J1-8
«W<1>
2...J~02
o~ ~ .
1-1..L..:3
~OO
~ : 0.1
o
~
o
.,...
1) 0.0
£ 1400
-
,.....
....
<1>
-
....
5 0.5
~
o
....
(,)
~
~ ~ 0.4
(1)0""
a:::2~
W en
~~.~ 0.3
028
U <1>
...JI-~
«W....
2 ~! 0.2
0-""
I-I..L..0
uoZ
« -
a::: 001
I..L.. ~ .
o
.,...
....
(,)
~ 0.0
~ 1400
-
- 75 -
Figure 2-9
EFFECT OF BED TEMPERATURE ON
CONVERSION OF INLET N TO NO (ALUNDUM BED)
U = 6 fps
02 (FI ue Gas) = 4i'o
Ho = 6 Inches
.
--.
Fuel N
.
-
.
.-
--t
.
250 ppm NO added
to fluidizing gas
1500
1600 1700
BED TEMPERATURE, OF
1800
Figure 2-10
EFFECT OF BED TEMPERA TURE ON
INLET N CONVERSION TO NO (CaS04 BED>
U = 6 fps
02 (FI ue Gas) = 4i'o
Ho = 6 Inches
Fuel N
.
250 ppm NO added
to fluidizing gas
1500
1600 1700
BED TEMPERATURE, of
1800
1900
1900
-------
1000
- 76 -
Figure 2-11
EFFECT OF 02 IN FLUE GAS ON
NO EMISSIONS (ALUNDUM BED)
-..
»
.a 900
--
S Normal i zed to
0.,
p.. 3/'0 ° 2
800
00
Z
0
1-4
00
00 700 Bed material = AI undum
1-4
~ (20 mesh)
~ T = 16000F
0
Z 600 U = 6 fps
(dp) coal = 300 }J Actual
500
o
900
~ 800
""tJ
'-'
E
a.
a. 700
..
(/)
z
a
~ 600
:2
w
~ 500
400
o
2
4 6
PERCENT 02 IN FLUE GAS
8
Figure 2-12
EFFECT OF 02 IN FLUE GAS ON
NO EMISSIONS (CaS04 BED)
U = 6 fps
T = 1600°F
~ed Material = CaS04 (-1000JL)
(dp) coal = 300,.,.
Normal ized
to 3/,002
Actual
2
4 6
PERCENT 02 IN FLUE GAS
8
10
10
-------
- 77 -
Actual NO emissions, using both Alundum and CaS04 as bed
materials, decreased as excess air (percent 02) was increased.
However,
when the emissions were normalized to a constant gas volume (3% 02)' the
NO emissions increased as the excess air was increased.
The NO formation
rate was thus increased by the higher average oxygen concentration in
the bed.
The NO emissions were consistently lower over the CaS04 bed, as
compared to the Alundum bed, which confirmed the previous observation
that greater reduction of NO was occurring over CaS04 compared to Alundum.
The difference in the NO emissions from the two beds, Alundum and CaS04' in-
creased as the excess air was decreased.
Also, the CO concentration in the
flue gas increased as excess air was decreased.
This result lent further
support to the postulate that NO and CO were reacting over CaS04 and that
the rate of this reaction would be increased by high CO concentrations.
To further examine the effect of ° level on NO emissions,
2
295 ppm of NO was added to the fluidizing gas in, experiments with the
CaS04 bed.
The flue gas NO concentrations following this addition are
given in Appendix 4.
The fractional conversion of fuel nitrogen to NO
(before addition) and the fractional conversion of added nitrogen
(as NO) to NO are shown in Figure 2-13.
The high reduction of the NO added
to the fluidizing gas again reflects the passage of this gas through the
highly reducing zone above the grid.
-------
0.5
~ 0.4
(/)0
a::::z
W
>0
ZI- 0.3
8z
-JI-
-------
- 79 -
3.2
NO-CO Reaction Studies
The observed effects on NO emissions produced by
varying the excess air level, the combustion bed temperature, and the
fluidizing bed material indicated that part of the NO formed by oxidation
of the nitrogen in the coal was being reduced in the bed--most probably
by the CO that was known to be present from analyses of the flue gases.
To confirm this, and to investigate the characteristics of the NO-CO reaction
and its catalysis by CaS04 and CaO, two series of experiments were made
in the fixed bed reactor system described in Section 1.2.
The first
of these was made using the 2.25 inch ID ceramic reactor, with different
concentration of NO and CO blended with nitrogen and passed through beds
of CaS04 (Drierite).
Results of this series of experiments are summarized in
Table 2-3.
Negligible reaction occurred when the simulated gas was
passed through the reactor containing only the alumina bed support.
Table 2-3
Differential Reactor Results of
NO-CO-CaS04 Reaction System
Reactor
Bed Material
Bed Weight
Gas Velocity
=
2.25 inch
CaS04
75 grams
2 fps
ID ceramic tube
=
=
Run Number 40-A 40-B 37-A 37-B 38-A 38-B
Bed Temperature, of 1300 1300 1500 1500 1700 1700
Gas Volume, CFM 0.99 0.99 0.90 0.90 0.81 0.81
Inlet Gas Composition
NO, ppm 2017 2017 1990 1990 2025 2025
CO, ppm*
N2 Bal. Bal. Bal. Bal. Bal. Bal.
Outlet Gas Composition
NO, ppm 2000 1959. 1952 1865 1980 1912
CO, ppm 1700 3425 1670 3670 1720 3180
N2 Bal. Bal. Bal. Bal. Bal. Bal.
* not meas ured
-------
- 80 -
From the results summarized in Table 2-3, reaction rates were
calculated and are shown in Figure 2-14 as a function of the reciprocal
absolute temperature.
The reaction rate is seen to exhibit an
Arrhenius type dependence on temperature.
The apparent activation energy
derived from these experimental results is 8.5 kcal g. mole and
6.5 kcal per g. mole for outlet concentrations of 3400 ppm and 1700 ppm
respectively.
From these very limited data, the NO-CO reaction rate
appears to depend on the square of the CO concentration (i.e., is 2nd
order in CO concentration) since the rates measured at 3400 ppm outlet
concentration averaged about four times those obtained at 1700 ppm.
Additional information relating to the order of the CO-NO
reaction was obtained in a second series of fixed bed experiments.
In
this series of experiments, the reaction of NO with CO was studied over
calcined limestone in the 1" reactor.
Three concentration levels of
NO and CO in N2 were studied - 500, 1000 and 2000 ppm.
The results
obtained in this series of experiments are shown in Table 2-4.
Table 2-4
Differential Reactor Results of
NO-CO-CaO Reaction System
Reactor - 3/4 inch ID Steel
Bed Material - 1359 Limestone
Bed Weight - 60 grams
Gas Velocity - 2 fps
Tube
Inlet NO Concen tra t ion, ppm 500 4 1000 . 2000
Inlet CO Concentration, ppm 1000 500 1000 2000 1000
Outlet NO Concentration, ppm 0 510 115 0 1150
Outlet CO Concentration, ppm 700 0 0 725 0
-------
60
50 -
40 -
30
.
20
~!u
-IU
S2iulO
_'e:..>
;.!:/) 9
(;)- 8
g 7
o 6
~Z 5
I
4
3
2
1
.80
- 81 -
FIGURE 2-14
EFFECT OF TEMPERATURE ON THE NO-CO REACTION RATE
.
Outlet CO = 3400 ppm
e.
Outlet CO = 1700 ppm
o
Bed Material = CaSO 4
Bed Weight = 75 grams
Gas Velocity = 2 fps
. 98 1. 0 1. 02 1. 04 1. 06 1. 0
-------
- 82 -
It can be seen that the limiting reactant was essentially completely
consumed in these experiments.
Also, the decreases in NO and CO
concentrations were on a one for one basis.
Coupled with the fact that
this reaction showed a second order dependence on the CO concentration,
these experiments indicated that the reaction between NO and CO proceeds
by equation (2-3) rather than (2-4).
2 CO + 2 NO
.
2 C02 + N2
(2-3)
CO + 2 NO
.
C02 + NZO
(2-4 )
3.2.1
Effect of Sub-Stoichiometric Combustion
Having shown that the CO-NO reaction can be important in
establishing the level of NO emissions obtained in the fluidized bed
combustor, additional experiments were made to determine the effect of
operating such a combustor with less than the stoichiometric air required
for complete coal combustion.
Under these conditions, NO emissions
were expected to be low because of the high CO concentrations, and
because oxidation of fuel nitrogen would be retarded.
The experiments that were made used a partially sulfated lime
bed and stoichiometric air levels ranging from about 68% to 130%.
The
NO emissions obtained as a function of the stoichiometric air level
are shown in Figure 2-15 and complete data for these experiments are
given in Appendix 4.
NO emissions are seen to fa11 rapidly below about
the 80% stoichiometric air level and became immeasurable at the 68% level.
-------
800
700
-
~ 600
"'C
-
E
c..
c..
.. 500
(/)
z
o
~ 400
~
UJ
~ 300
200
100
o
60
FIGURE 2-15
NO EMISSIONS AS A FUNCTION
OF PERCENTAGE STOICHIOMETRIC AIR
.
.
Unnormalized
00
w
. ALUNDUM BED 6 fps (see Appendix 3)
. 20%S ULFA TED LIME BED 3 fps
(see Appendix 3)
I
I
70
90 100 110 120 130
PERCENTAGE STOICHIOMETRIC AIR
140
150
80
-------
- 84 -
The sub-stoichiometric operations essentially simulated the
first stage of a two-stage combustor.
The combustible losses in these
operations were high, with CO emissions being greater than 6000 ppm
(range of instrument) at all conditions below 90% stoichiometric
air addition.
Continuous measurement with the DuPont 461 Analyzer showed
negligible N02 emissions over the entire range of excess air
levels examined.
The air level at which NO emissions were observed to decrease
rapidly should depend upon the combustion efficiency attained, which,
in turn, is dependent upon operating variables such as bed temperature
and superficial gas velocity.
The combustion efficiencies for the data
shown in Figure 2-15 were generally below that possible for complete
utilization of the available oxygen.
Higher combustion efficiencies
would probably shift the entire
curve
shown in this figure to the right.
3.2.2
Effect of H20 on the
NO-CO Reaction
A final series of experiments on the NO-CO reaction was made
to determine the effect of water vapor On the reaction.
The 2.5 inch
electrically heated reactor was used for these experiments.
NO, CO and
N2 cylinder gases were blended and stearn was added to give the desired
reactor inlet gas concentration.
The results of these experiments are
given in Table 2-5.
-------
Table 2-5
Results of Study of NO-CO-H20 Reaction System
Run NO. 37-A 37-B 37-C 38-A 38-B 38-C 39-A 39-B 41-A 41-B
Bed Material caS04 CaS04 CaS04 CaS04 CaS04 CaS04 Alumina Alumina None None
Bed Temp, of 1500 1500 1500 1700 1700 1700 1500 1500 1500 1500
Inlet Gas Volume, CFM 0.90 0.935 0.97 0.81 0.84 0.87 0.90 0.97 0.90 0.97
I~let Gas Composition
NO, ppm (dry) 1990 1990 1990 2025 2025 2025 2010 2010 2035 2035
CO., ppm (dry) 4000 4000 4000 4000 4000 4000 4000 4000 4000 4000
H20, % 0 3.7 7.2 0 3.9 7.3 0 7.2 0 7.2
N2 ba1 ba1 ba1 ba1 ba1 ba1 bal ba1 ba1 ba1
Outlet Gas Composition (dry)
NO,ppm (dry) 1915 1730 1745 1900 1575 1575 1990 1750 2015 1725 I
CO, ppm (dry)* ex>
H20, %** VT
N2 ba1 bal ba1 ba1 ba1 ba1 bal ba1 . ba 1 bal
Fractiona 1 Conversion of NO, % 3.8 13.0 12.3 6.2 22.2 22.2 1.0 13.0 0.5 15.2
* Not measured.
** Not measured, gas passed through 350F refrigerator before analysis.
-------
- 86 -
The addition
of a few percent steam (3.7%) is seen to enhance
the CO reduction reaction.
Higher concentrations of steam (i.e., 7.2%)
do not increase the NO conversion and may in fact decrease it slightly.
Runs 38 A, Band C show the effect of reactor temperature on the NO
conversion in the presence of the water.
The increase in temperature
from 1500 to l700°F about doubled the fractional conversion of NO with
the steam present but had very little effect without the steam.
This
confirms the low activation energy noted in Section 3.2 for the NO-CO
reaction when it occurs in the absence of steam.
Runs 39 and 41 show that the addition of steam to the system
minimizes the effect and the need for an active solid to promote 'the
NO-CO reaction.
Whereas in the absence of steam (Run 39A and 4lA vs 37A) ,
the alumina supports and the empty reactor gave less conversion than
obtained with the CaS04' about the same NO conversion level occurred
with the empty reactor and the alumina supports when steam was present.
These results would indicate that the accelerating effect of steam on
the rate of NO conversion is probably a gas phase reaction, or a reaction
catalyzed by the wall of the ceramic reactor itself.
~
I
J
-------
- 87 -
3.3 NO-SOZ-CaO Reaction System
Typical emissions of NO and 502 obtained in the Esso FBC are
shown in Figure
2-16.
The change in 502 emissions reflect the batch nature
of this reaction system; the bed was batch charged with lime, and coal
was continuously fed.
Initially, the reactivity of the lime was high
enough that essentially no 502 appeared in the flue gas.
As the run
progressed, the lime reacted with S02 to form CaS04' and eventually
the CaS04 concentration became great enough that not all of the 502 was
captured by the lime and 802 appeared in the flue gas.
Figure 2-16
TYPICAL EMISSIONS WITH LIMESTONE BED
1250
........ NO
>. X
M 1000
o
----
8
p..
p..
T = 1600°F
U = 3 fp s
Ho = 12 in.
Excess Air = 1010
(2P) stone = 930J!
(dp) coal = 200J!
750
...
c::
o
.....
en
en
.....
8
~
500
en
:;:j
o
Q)
en
CI1
c.=>
250
o
o
1
2
3
4
5
Run Time, hours
-------
- 88 -
In order to obtain further information on this reaction system,
a differential reactor study was made.
The fixed bed reactor system
shown in Figure Z-4 was used in this study.
Premixed cylinders of NO in
NZ (at different concentrations) were used to assure uniformity of inlet
concentration.
Pure SOZ was added to the inlet gas as desired.
The
dilution effect of the added SOZ was negligible.
Complete results of
These experiments are given in Appendix 6.
These experiments again
showed that no reaction occurred between NO and S02 over CaS04 or Alundum.
A reaction did occur over partially-sulfated lime.
The apparent order
and the temperature dependence for this reaction are given in Sections
3.3.2 and 3.3.3.
The characteristics of the partially-sulfated lime
used are given in Appendix 7.
3.3.1
Examination of Rate Controlling Mechanism
in the NO-S02-CaO Reaction System
Fixed bed reaction studies were made in the 2-1/2 inch externally
heated unit to determine if gas film diffusion was the controlling factor
in the NO-S02-CaO reaction system.
bed material)/(inlet gas volume) ratio was maintained constant at different
In these experiments, the (weight of
bed weights.
In the region in which gas film diffusion is Q£! controlling,
the fractional conversion of the reactants should be constant in such
a series of experiments (l).
As shown in Table 2-6, the fractional
conversion
of inlet NO was essentially constant in the experiments conducted,
indicating that gas film diffusion was not controlling.
The higher
conversion obtained with the smaller bed probably reflects the fact that
the upper surface of this bed was not restrained, allowing for some
expansion.
The temperature within the deeper fixed bed, measured at two
axial positions and two radial positions, was 1475 + 50°F.
-------
- 89 -
Table 2-6
Conversion of NO at Constant W/F
Indicates Film Diffusion Not ControllinR
Bed Temperature = l500°F
Inlet NO = 580 ppm
Inlet = 1900 ppm
Bed Material = 16.6% Sulfated
Lime
Run
Number W. Grams F. cfm W/F. grams/CFM ..1L-
10 150 0.90 167 0.43
13 300 1.80 167 0.37
As shown in Ffgure 2-16, NO emissions from the Esso FBS,were
typically above 1000 ppm at the start of the run, but decreased as the
run progressed.
Since the only two factors that changed with time were
the increase in S02 concentration and the sulfation level of the lime,
it appeared that one or both of these changes were causing the downward
trend observed in NO emissions.
With the objective of pinpointing the cause for the downward
trend in NO emissions, a number of experiments were made with different
bed materials in the 2.5" fixed bed reactor shown in Figure 2.4.
Gases
containing NO and S02 diluted with nitrogen were used for these experiments.
To determine the extent of reaction, if any, occurring in the gas phase,
one experiment was carried out without any bed material present in the
reactor.
A summary of the results of these experiments is given in Table 2-7.
-------
- 90 -
Table 2-7
Effect of Bed Material
on the NO-S02 Reaction
Reactor Temperature - l600°F; Superficial Gas Velocity - 2 fps;
Settled Bed Height - Approximately 6 In.
NOx Cone., ppm
Before S02 After S02
Intro. Intro.
SOz Cone., ppm
Inlet Outlet
Bed M~terial
Transport Gas
1. Gas Phase
2a) Partially-Sulfated Lime*
b) Partially-Sulfated Lime*
3. Alundum
4. CaS04 (Anhydride)
* From the Esso FBC.
900
840
830
820
860
- N
2
900
440
180
820
860
1290
785
1510
1000
670
1290
300
480
1000
670
These results showed that No and SO did not .
2. react ~n the gas phase,
over Alundum or over CaS04 at the conditions examined.
However, a reaction
did Occur with partially-sulfated lime.
The rate of this reaction increased
with increasing S02 concentration.
reaction system involved CaO.
Therefore, it appeared that the NO-SO
Z
3.3.2
Apparent Reaction Order
with Respect to NO
In order to determine the apparent order of the NO-S02 reaction,
additional fixed bed reaction studies were carried out with the 2-1/2 inch
reactor.
premixed cylinders of NO in NZ at concentrations of 2110 ppm,
1035 ppm and 485 ppm were used for these studies.
The concentration of
the cylinder gas was analyzed, and then flow to the hot reactor initiated.
The reactor outlet NO concentration was'measured.
On attaining a steady
state condition, S02 was blended to the inlet gas at the desired concentration
and the outlet NO and S02 concentrations were monitored.
The apparent order of the reaction with respect to NO was
determined by assuming the rate equation to be of the form given in
equation
(2-5) .
- r
NO
- -
1 d(NO) = k(NO)n (SOz)m
W dt
(2-5)
-------
- 91 -
If the 802 concentration is assumed to be greatly in excess of the NO
concentration and therefore to be constant, the apparent order of the
reaction system with respect to NO can be determined by plotting In (- ~ d~:O) )
versus
In (NO).
The experimental results plotted this way are shown in
Figure 2-17.
The 802 concentrations shown are the arithmetic averages of
the inlet and outlet 802 concentrations.
(i.e., the slopes of the curvest was 0.53 at an average 802 concentration
The apparent order of the reaction
of about 550 ppm and 0.67 at an average 802 concentration of about 1650 ppm.
An order of approximately 0.5 is consistent with a Langmuir-
Hischelwood mechanism (~}
for a single reactant absorbing on an active
site.
The rate equation for this mechanism is given by:
- r
NO
= kKNO
1 + KNO
(2-6 )
This equation can be approximated by:
n
- rNO + k(NO)
where
O
-------
VI~
O~
E 0
Or:::
..Q .-
- E
-
U"'I
o
r-t
'-0
~
I
- 92 -
The rate data obtained are thus consistent with an overall
reaction scheme involving NO, S02 and CaO as represented by equations
(2-11) and (2-12) 0
.
CaS03
(2-11)
CaO + S02
2CaS03 + 2NO
.
(2-12)
2CaS04 + N2
where it is assumed that CaO and S02 react to form CaS03' which then
acts as an active site for converting the NO to N2°
100
10
1
100
Figure 2-17
APPARENT REACTION ORDER WITH RESPECT
TO NO OF NO -SO -CaO REACTION SYS TEM
X X
Reactor = 2.25 in ID Ceramic Tube
Bed Material = 16. 6 % Sulfated Lime
Bed Weight = 75 grams
Gas Rate = 0.90 ft.3/min.
Gas Components = NO, S02' N2
Bed Temperature = 1500°F
\
(S02) = 1650
SLOPE = 0.67
(S02) = 550
SLOPE = 0.53
1000
C NO' ppm
10,000
-------
- 93 -
3.3.3
Temperature Dependence of
NO-S02-CaO System
The temperature dependence of the NO-S02-CaO reaction system
was next examined in the 2-1/2 inch fixed bed unit over the temperature
range from 1400°F to l600°F.
The results of these experiments are given
in Appendix 6.
As shown in Figure 2-18, this reaction system demonstrated
a negative temperature dependence in the range of l400°F to l600°F.
This negative temperature dependence is also consistent with
the reaction mechanism represented by equations (4-10) and (4-11).
Since
CaS03 is probably the reactive intermediate in this reaction sequence,
higher CaS03 concentrations should lead to a greater rate of NO disappearance.
The 802 concentration in equilibrium with CaS03 and CaO is given in
Figure 2-19 as a function of temperature.
CaO + S02 .
t
CaS03
(2-13)
For the inlet 802 concentration of 700 ppm used in this study,
the equilibrium temperature from Figure 2-19 is l430°F.
This means that
the CaS03 is thermodynamically stable at all temperatures below l430°F
and unstable above l430°F.
Therefore, ~ rapidly accelerating rate with de-
creasing temperature for the CaO-802-NO reaction system beginning at about
1400° F, would be expected.
-------
11'1 U
CIJ CIJ
011'1
E I
. U
C'lU
-
en
a
r-4
'-'
a
r!,Z
I
«
60
50
40
35
30
25
Figure 2-18
TEMPERATURE DEPENDENCE
OF NO-S02-CaO REACTION SYS TEM
20
15
10
9
8
7
6
5
4
0.86
Inlet Gas Composition
NO = 485 ppm
S02 = 700 ppm
N2 = Bal.
Gas Residence Time = 0.04 sec.
Bed Material = 16.6% Su !fated Lime
1.0
~
* Rate measured 15
minutes after gas
flow started to
reactor.
0.88
0.90
0.92 0.94
1000/T, OK
0.96
0.98
1.0
-------
- 95 -
Figure 2-19
502 EQUILIBRIUM OVER Ca503
100,000
80,000
60,000
40,000
20,000
REACTION:
CaO + 502 ==;;
Ca503
10,000
8,000
6,000
8 4,000
0..
0.
N
o
V) 2,000
a..
------
1,000 I
800 I
-----
600
400 I I
I
I
200 I I
I I
100 I I
80
60 I I
40
800 900 1000 1100 1200 1300 1400 1500
TEM PERA TURE, OK
1160 1340 1520 1700
TEMPERATURE, of
-------
-7' -
3.3.4
Miscellaneous NO-SOZ-CO-CaO
Reaction Studies
Additional NO-SOZ-CaO reaction experiments were made using
the 1 inch fixed bed reactor system described in Section 1.Z.Z.
Gases
containing 1000 ppm NO, and 500, 1000 and 2000 ppm S02 in NZ were used
in these experiments carried out at l600°F.
Results are summarized
in Table Z-8.
Table 2-8
Additional Fixed Bed
NO-SOZ-CaO Reaction Studies
Reactor = 3/4 inch ID
Bed Material = -16 to +18 mesh CaO
Bed Weight = 60 grams
Gas Velocity = 2 fps
Reaction Temperature = l600°F
Reactor Inlet Concentration, ppm
NO 4 1000 .
SOZ 500 1000 ZOOO
Reactor Outlet Concentration, ppm
NO 650 550 380
80Z 10-Z5 15-70 40-Zl0
I '
-------
-"-
A comparison of these results with those obtained for the
CO-NO reaction carried out in the same unit and under the same conditions
(see Table 2-8 in this section) shows that the reaction of NO with CO is
faster than is its reaction with S02' at least in the absence of moisture.
This can be seen by comparing corresponding outlet NO levels at the
1000 ppm inlet NO concentration for the tests made with the 500 and 2000
ppm concentrations of CO and 502'
In the final studies made with the 1 inch fixed bed reactor
system, evidence that CaS may playa role in the reduction of NO was
obtained by reacting CaO sequentially with S02' then CO, and then NO.
The results from this experiment are shown in Table 2-9.
The NO emissions
observed after subjecting the CaO to 2000 ppm of S02 followed by 2000 ppm
of CO remained below the inlet NO concentration of 1000 ppm for over two
hours after the flow of CO had been discontinued.
This may reflect a
reduction of NO by the CaS formed from the reaction of CaS03 with CO
as shown below:
CaO + S02
. CaS03 (2-14)
. CaS + 3 C02 (2-15)
. 3 (2-16)
CaS03 + 2 N2
CaS03 + 3 CO
CaS + 3 NO
-------
Table 2-9
NO-CO-S02-CaO Reaction Studies.
Operating Conditions - Same as Table 2-8
Step 11 Time Period. Min. Inlet Gas. ppm. Outlet Gas. ppm. Remarks
1 0-60. 2000. 8°2 0-225. 8°2 Gradual increase over period.
2 60-75. 2000. CO O. CO Rapid increase from 0.-1150 ppm CO
75-120 2000. 00 1150.-1500. CO During this period 802-0.
3 j 120-160. 10pO. NO O. NO Rapid increase from 0.-470. ppm NO
I
" 160-240. 1000. NO 470.-800. NO During this period 802-- o. for 25 min.
, 0.-250. For 25 min
t
I . 250.-100. For 15 min
. 100.-0. For 55 min
, J~
- CO - O.
\D
00
-------
- 99 -
3.4
Examination of Two Stage Combustor
Operation for NO Emissions Control
~
The experimental data obtained in this program indicate that
at least two reactions systems are important in determining the NO
x
emissions from a fluidized bed combustor.
One involves the reaction
of NO, S02 and CaO.
The experimental data indicate that this reaction
proceeds through a CaS03 intermediate.
competing reaction, given by equation (2-17), can occur which limits the
In the presence of oxygen, a
extent of NO reduction.
2 CaS03 + 02
~
2 CaS04
(2-17)
Therefore, to maximize NO reduction by reaction with S02 and CaO, it
would be desirable to minimize the oxygen concentration in the FBC.
The
second reaction which can occur in fluidized bed combustors involves
the CaO catalyzed reduction of NO by CO.
Thus, it would be highly
desirable to have reducing zones in the fluidized bed combustor to maximize
NO
x
reduction.
Both of these NO reactions would be accelerated by operating
the fluidized bed combustor with staged addition of air.
In this mode of
operation, part of the combustion air would be supplied through the grid
of the combustor and the remainder of the combustion air would be supplied
at some level higher-up but still in the bed.
By supplying substoichiometric
air (based on the coal rate) through the grid, the lower part of the bed
would necessarily be deficient in oxygen; high CO concentrations wouid
be present.
The second stage air would serve to burn out the CO and
generate S02.
Consequently, both the NO-S02-CaO and the NO-CO-CaO reduction
systems would be operative in a two stage system.
-------
- 100 -
.
In order to verify the effectiveness of two stage combustion
as a NO control technique, the Esso FBC was modified to allow its operation
x
as a two stage combustor.
This modification, shown in Figure 2-3, consisted
of placing a 1/4 inch stainless steel tube horizontally through the bed,
six inches above the grid.
Holes were drilled around the upper circumference
of this tube so that it acted essentially as a second air distributor.
The results of operating the Esso FBC as a two stage combustor
are shown in Figure 2-20.
As the ratio of the second stage air to the
first stage percentage air was increased (i.e., as the first stage became
more reducing), NO emissions decreased.
Only two bed temperatures were
examined, l600°F and l750°F, with the effect of staging being more
pronounced at the higher temperature.
The lowest NO emissions observed in these staging runs was
200 ppm, which was achieved at l600°F and 43 percent stoichiometric air
to the first stage.
Although these highly reducing first stage conditions
may not be feasible for commercial operation, the potential for NO
x
reduction by two stage combustion is clearly apparent.
It is significant
that CO emissions in all staging operations remained relatively low, at
about 400 to 600 ppm.
-------
- 101 -
FIGURE 2-20
STAGED FBC RESULTS
800
700
Bed Material = CaSO 4
K 600
a.
Total Air = 110CYo of Stoichiometric
T = 1600°F
..
U) 500
2
o
V> 400
U)
~ 300
o
2 200
Total Air = 130CYo of Stoichiometric
T = 1750°F
100
o
o
0.5
1.0
SECOND STAGE AIR
FIRST STAGE AIR
1.5
-------
- 102 -
REFERENCES
1.
Hammons, G., and Skopp, A., A Regenerative Limestone Process for Fluidized
Bed Coal Combustion and Desulfurization, Final Report to Air pollution
Control Office under Contract CPA 70-19, February (1971).
2.
Archer, D. H., et. al., Evaluation of the Fluidized Bed Combustion Process,
Fifteenth Monthly Progress Report submitted to Air pollution Control Office
under Contract CPA 70-9, March (1971).
3.
Craig, J. W. T., et. al., Study of Chemically Active Fluid Bed Gasifier
for Reduction of Sulfur Oxide Emissions, Interim Report to Air pollution
Control Office under Contract CPA 70-46, February (1971).
4.
Jonke, A. A., et. al., Reduction of Atmospheric pollution by the Application
of Fluidized-Bed Combustion, Annual Report to Air pollution Control Office
ANL/ES-CEN-1002, June (1970).
5.
Levenspiel, V., Chemical Reaction Engineering, John Wiley and Sons, New
york, (1966).
6.
Laidler, K. J., Chemical Kinetics, McGraw Hill Book Company, New york (1966).
7.
White, S. B., et. a1., . Chemical Equilibrium in Complex MixtuTes,
Rand
Corp., Bulletin p-1059.
-------
- 103 -
NOMENCLATURE
atm.
Pressure in atmospheres
CMTU
Esso's Cold Model Test Unit
dp
Weight Median Particle Diameter, microns
F
Inlet Gas Rate, cubic feet per minute
FBB
Fluidized Bed Boiler
FBC
Fluidized Bed Combustor
FBCR
Fluidized Bed Combustor-Regenerator (Miniplant)
fps
Feet Per Second
h
Settled Bed Height, feet
Ho
Settled Bed Height, inches
H-
S
Average Bed Slug Height, feet
k
Rate Constant
K
Absorption Equilibrium Constant
Kp
Equilibrium Constant
n, m
Reaction Orders
P
Fluidized Bed Pressure in psia
ppm
Parts Per Million
-rNO
Rate of Disappearance of NO
(S02)
Arithmetic Average S02 Concentration
Standard Cubic Feet Per Hour
SCFH
SCFM
Standard Cubic Feet Per Minute
T
Bed Temperature, of
U
Superficial Gas Velocity, feet per second
I.l
Microns
v
Incipient Fluidization Velocity, feet per second
W
Bed Weight, grams
WTM
Wet Test Meter
-------
- 104 -
APPENDIX 1
PREDICTED S02 CONCENTRATION
FOR PRESSURIZED REGENERATION
The regenerator internal diameter depends on the S02 concentration
achieved in the regenerator off-gas.
Although pressurized regeneration
experiments have not yet been performed, the theoretical S02 concentration
that would be attained in the regenerator off-gas at thermodynamic equilibrium
has been calculated for a number of different reducing gas compositions at
1 and 10 atmospheres.
A computer program developed by the Rand Corporation
(2), in which the total free energy of the reacting system is minimized, was
used for these calculations.
The predicted equilibrium regenerator off-gas
concentrations are given in the table below for two different assumptions
regarding the thermodynamic state of the solid phase.
The first set of pre-
dictions assumes unit activity for each of the solid chemical species, while
the second set assumes the activity of each of the solid components is
proportial to its molar concentration in the total solid phase.
The actual
activities of the solids probably fall somewhere:between these extremes.
PREDIcrED EQUILIBRIUM RESULTS FOR
SULFATED LIME REGENERATION AT 1 AND 10 ATMS.
I
1 Atmosphere 10 Atmosphere
Reducing Gas M % S02 Moles CaS04/l00 Moles CO M % S02 Moles CaS04/l00 Moles CO
% CO % C02 In Regen. .converted to: In Regen. converted to:
in N2 Off-Gas CaO CaS Off-Gas CaO CaS
10 10 9.1 99.6 8.5 94.2
10 20 9.0 99.4 8.4 92.6
20 20 16.6 99.3 9.5 53.7 10.0
Activity of Solids c(. to Concent rat ion
10 10 7.9 84.9 2.4 4.9 46.8 9.8
10 20 7.9 84.4 2.1 5.0 45.4 9.3
20 20 13.5 76.7 4.1 7.8 35.2 11. 9
-------
- 105 -
APPENDIX 2
CALCINING PROCEDURE
An electrical oven is used to calcine the limestone in this
study.
The oven is initially brought to 1700°F.
Two thousand grams of
the limestone are then charged to the oven, resulting in a settled bed
of limestone approximately 1-1/2 inches in depth.
A constant nitrogen
bleed of four liters/minute is maintained above the bed.
The system is
maintained at 1700°F for seven hours, at which time the electrical
heat is sh ut off.
The oven is allowed to cool for nine hours before
removing the calcined stone at 500°F.
The nitrogen sweep gas is maintained
during the cooling-down period.
-------
~13
::r::
5 12
~
g.
ell
Q
~1O
~
fI)
~ 9
r-.
P4
~ 8
Q)
co
16
15
14
7
6
5
4
3
2
1
o
APPENDIX 3A
MINIMUM FLUIDIZING VELOCITY - 20% SULFATED LIME N-1359
Bed Material - 20% Sulfated N-1359 Lime (930 f)
Bed Weight - 510 grams
Fluidizing Gas = Air, 70°F, 1 Atm.
Bulk Density, Lime = 0.975 gram/cc (loose) .
0.5
1.0
1.5
Superficial Gas Velocity, FPS
o
Average
o
....
o
0-
2.0
2.5
-------
o
N
::t::
8 25
C,)
0..
o
M
Q
Q)
M
~ 20
rI)
Q)
M
p..
'tj
Q)
~
45
40
35
30
15
10
5
o
- 107 -
APPENDIX 3B
MINIMUM FLUIDIZING VELOCITY - ALUNDUM
Bed Material - Alundum (841]1)
Bed Weight = 1315 grams
Fluidizing Gas = Air, 70°F
Bulk Density = 2 grams/cc (loose)
Slugging
Onset
o
2
3 4
Superficial Gas Velocity, fps
7
5
1
6
-------
APPENDIX 4
Summary of Operations - Alundum and CaS04 Beds
Coal -
Coal dp -
Superficial Gas Velocity =
Settled Bed height -
Run Number
Run Length, Hrs.
Bed Material
Bed Material, dp
Bed Temperature, of
Coal Feed Rate, lbs/hr
Air Rate, CFM
Percentage Stoich. Air
Carbon Loss in Fly Ash, %
Flue Gas Analysis (dry)
C02' %
02' %
N2' %
CO, ppm
S02' ppm
NO, ppm
NO Added to Fluidizing
Gas, ppm
Flue NO After Addition, ppm
Material Balance, (Out/In) X100
Carbon
Oxygen
S ulf ur
Bit. A
350 Jl
6 fps
6 In.
32A
32B
1.1
<
<
<
1.56
4.55
130.5
12.6
1.3
Alundum
840 r
1600
1.85
4.55
110.0
13.4
10.8
8.0
11. 8
6.0
450
2050
650
550
2325
720
93.6
102.0
114.0
91.5
97.2
111.0
32C
1.3
2.03
4.55
101.0
13.1
14.3
4.0
1200
2625
750
95.6
103.0
113.0
32D
1.3
,.
~
..
2.50
4.55
81.6
14.3
16.0
1.0
to-'
a
(XJ
6000+
3000+
830
93.8
100.0
108.0
-------
APPENDIX 4
Summary of Ope rations - Alundum and CaS04 Beds (Con I t)
Run Number 33A 33B 33C 33D 34A 34B 34C 34D
Run Length 1 1 1 1 1 1 1 1
Bed Material < Alundum r " CaS04 ~
Bed Material, dp < 840 E :>- c: 1000 ..,.
Bed Temperature, of 1500 1600 1700 1800 ('" 1600 >
Coal Feed Rate, 1bs/hr 2.20 2.02 1. 75 1.63 1.57 1. 83 2.07
Air Rate, cm 4.78 4.55 4.33 4.14 < 4.55 :::>
Percentage Stoich. Air 97.3 101 111 98.5 130.5 111 . 8 98.3 81. 3
Carbon Loss in Fly Ash, % 17.6 14.9 13.4 10.8 12.3 12.1 11.3 13.2
Flue Gas Analysis, (dry)
C02 13.6 14.4 13.7 14.0 10.5 12.0 14.1 . 16.2
02 4.0 4.0 4.0 4.0 8.0 6.0 4.0 1.0
N2
CO 5200 1800 570 250 350 570 820 6000+ ~
o
802 2625 2525 2325 2250 2100 2500 2800 3000+ '-P
NO 665 750 770 745 565 620 650 650
NO Added to Fluidizing
Gas, ppm 254 253 252 253 295 293 295 295
Flue NO After Addition, ppm 705 790 820 790 615 660 680 660
Material Balance, (Out/In) X100
Carbon 96.3 100.0 101.0 102.0 91;8 89.7 90.6 95.4
Oxygen 102.0 104.0 98.8 102.3 98.8 99.3 103.5 98.2
Sulfur 116.0 113.0 112.0 108.0 117.0 119.0 117.0
-------
APPENDIX 4
Summary of Operations - Alundum and CaS04 Beds (Con' t)
Run Number 35A 35B 35C 35D 35E
Run Length, Hrs. 1 1 1 1 1
Bed Material ~ CaS04 >
Bed Material, dp < 1000}l >
Bed Temperature, of 1400 1500 1600 1700 1800
Coal Feed Rate, 1bs/hr 2.39 2.13 1.99 1. 76 1. 80
Air Rate, CFM 5.03 4.78 4.55 4.33 4.14
Percentage Stoich. Air 94.2 100.1 103.0 110.0 103.0
Carbon Loss in Fly Ash, % 19.0 14.8 13.5 12.6 10.1
Flue Gas Analysis, (dry)
C02 13.2 13.0 14.0 14.2 14.4 ~
02 4.7 4.5 4.4 4.0 4.0 ~
o
N2
CO 6000+ 4650 2250 500 250
5°2 2700 2575 2500 2475 2700
NO 300 540 610 640 680
NO Added to Fluidizing
Gas, ppm 252 252 253 253 253
Flue NO After Addition, ppm 310 560 630 670 725
Material Balance, (Out/In) X100
Carbon 96.3 92.6 97.8 103.4 96.3
Oxygen 98.3 98.3 102.3 102.3 102.2
Sulfur 116.0 114.0 107 .0 119.0 117.0
-------
- 111 -
APPENDIX 4
Summary of Operations - Alundum and CaS04 Beds (Con't)
Run Number 16A 16B
Research Notebook 514 514
Bed Material < (\120% Sulfated
Lime N-1359
Bed Temperature, of 1600 1600
Coal Rate, 1b s /h r 1.14 1.50
Ai r Rate, CFM 2.27 2.27
Percentage Stoichiometric Air 88 68
Superficial Gas Velocity, FPS 3 3
Flue Gas Analysis (Dry)
03' % 1.0 0
C , pprn < 6000+
NO, pprn 780 0
16C
>
1600
1. 20
2.27
83
3
0.5
>
715
-------
- 112 -
APPENDIX 5
Sample Calculation of Fractional Conversions to NO
Fuel Nitrogen
Nitrogen content of coal = 1.4 weight percent
Run 35E
NO emissions for complete conversion of fuel
nitrogen to NO
Avg coal rate = 1. 80 1bs /hr
combustion eff. = 90%
Air rate = 4.14 ft3/min
1. 80 ~~s coal
x
0.90 x 0.014 1bs N
lbs coal
x
lb mole
14 1bs N
x
379" ft3
lb mole
x
min
4.14 ft3
x
hr
60 min
= 2470 ppm NO (dry)
where negligible gas expansion is assumed on a dry basis
F . 1 . f f 1 N to NO -- 680
ract10na converS1on 0 ue = 0.275
2470
where outlet NO = 680 ppm (dry)
NO added to fluidizing gas
Inlet NO concentration (based on total gas flow
through the reactor) = 253 ppm
Increase in Flue Gas NO concentration = 45 ppm
45
Fractional conversion of added N to NO = 253 = 0.178
(Fractional retention of added NO)
-------
. APPENDIX 6
Summary of Studies of NO - S02 Reaction System
Rtm Number 3A 3B 3C 4 9A 9B 9C 9D
Research Notebook 514 514 514 514 514 514 514 514
Bed Material Altmdum Altmdum Altmdum caS°4 CaS04 CaS04 CaS04 CaS04
Bed Weight, grams 300 300 300 300 150 150 150 150
Bed Temperature, of 1600 1600 1600 1600 1600 1600 1500 1500
Bed Height, inches 2.3 2.3 2.3 4.7 2.3 2.3 2.3 2.3
Gas Flow Rate, CFM. 0.91 0.91 0.91 0.91 0.91 0.91 0.91 0.91
Superficial Gas Velocity, fps 1.42 1.42 1.42 1.42 1.42 1.42 1. 35 1. 35
Inlet Gas Composition
NO, ppm 856 856 856 859 1025 992 1030 1019
S02, ppm 890 1895 1850 650 750 675 825
N2' % Bal. Bal. 79. Bal. Bal. 79. Bal. 79.
02' % 21. 21. 21.
Outlet Gas Composition ......
NO, ppm 848 860 820 848 1015 954 1035 954 ......
w
S02' ppm 875 1860 1850 650 750 675 700
N02' ppm 40 50 55
N2' % Bal. Bal. 79. Bal. Bal. 79. Bal. 79.
02' % 21. 21. 21.
-------
APPENDIX 6
S uunna ry of Studies of NO - S02 Reaction System
Run Number 1QA lOA lOB lOB 11A 11A lIB 11B
Research Notebook 514 514 514 514
Bed Material Lime * Lime* Lime* Lime*
Bed Weight, grams 150 150 75 75
Bed Temperature, of 1500 1500 1500 1500
Bed Height, inches 2.3 2.3 1.2 1.2
Gas Flow Rate, CFM 0.91 0.91 0.91 0.91
Superficial Gas Velocity, fps 1. 35 1. 35 1. 35 1. 35
Inlet Gas Composition
NO, ppm 585 585 1015 1015
S02' ppm 670 1750 670 1975
N2' % Bal. Bal. Bal. Bal.
02' %
Outlet Gas Composition
Time After Introduction .....
......
of Gas to Reactor, min. In . * * 15 In. 15 In. 15 In. 15 ~
NO, ppm 560 535 400 325 933 938 850 870
5°2' ppm 370 390 1265 1410 340 370 1180 1460
N02' ppm
N2' % Bal. Bal. Bal. Bal. Bal. Bal. Bal. Bal.
°2' %
* 16.6 percent sulfated lime N-1359 removed from Esso fluid bed
** bed combustor at 90 percent S02 removal.
Initial, approximately 2 minutes.
-------
APPENDIX 6
Summa ry of Studies of NO - S02 Reaction System
Reactor = 2.25 In. ID Ceramic Tube (Elect rically-heated)
Run Number 12A 12A 12B 12B 13A 13A 14A 14A
Research Notebook 514 514 514 514
Bed Material Lime* Lime* Lime* Lime*
Bed Weight, grams 75 75 300 75
Be d Tempe rat ure, 0 F 1500 1500 1500 1500
Bed Height, inches 1.2 1.2 4.7 1.2
Gas Flow Rate, CFM 0.91 0.91 1. 82 0.91
Superficial Gas Velocity, fps 1. 35 1. 35 2.70 1. 35
Inlet Gas Composition .....
NO, ppm 2110 2110 585 485 .....
V1
S02' ppm 700 1900 1950 700
N2' % Bal. Bal. Bal. Bal.
°2' %
Outlet Gas Composition
Time after Introduction
of Gas to Reactor, min. In. 15 In. 15 In. 15 In. 15
NO, ppm 2035 1990 1885 1835 370 382 435 460
S02' ppm 400 400 1350 1400 1590 1710 480 460
N02' ppm Bal. Bal. Bal. Bal.
N2' % Bal. Bal. Bal. Bal.
02' %
-------
APPENDIX n
Summary of Studies of NO - SO Reaction System
2
Reactor = 2.25 In. ID Cerami c Tube (E1e ct rically-heated)
Run Numbe r 14B 14B 18A 18A 18B 18B 19A 19A
Research Notebook 514 514 514 514
Bed Material Lime* Lime* Lime* Lime*
Bed Weight, grams 75 75 75 75
Bed Temperature, of 1500 1600 1600 1500
Bed Height, inches 1.2 1.2 1.2 1.2
Gas Flow Rate, CFM 0.91 0.90 0.90 0.945
Superficial Gas Velocity, fps 1. 35 1.40 1.40 1. 40
Inlet Gas Composition
NO, ppm 485 485 485 485
S02' ppm 1900 700 1900 700 ~
~
N2' % Ba1. Ba1. Ba1. Ba1. 0'\
02' %
Outlet Gas Composition
Time after Introduction
of Gas to Reactor In. 15 In. 15 In. 15 In. 15
NO, ppm 415 363 452 458 300 275 405 455
502' ppm 1210 1400 420 420 1400 1400 420 400
N02' ppm Ba1. Ba1. Ba1. Ba1. Bal. Ba1. Ba1.
N2' % Ba1.
02' %
-------
- 111 -
APPENDIX 6
Summary of Studies of NO - S02 Reaction System
Reactor = 2.25 inch ID ceramic tube (Electrically-heated)
RlU1 Nwnber
19B
19B
Bed Material
Bed Weight, grams
Bed Temperature, of
Bed Height, inches
Gas Flow Rate, CFM
Superficial Gas Velocity, fps
Inlet Gas Composition
NO, ppm
S02' ppm
N2' %
02' %
Outlet Gas Composition
Time after flow Started
to Reactor
NO, ppm
S02' ppm
N02' ppm
N2' %
02' %
Lime *
75
1400
1.2
1.0
1.40
485
700
Bal.
In. 15
382 377 .
400 400
-~~
Ba1. Ba1.
-------
- 118 -
APPENDIX 7
Method of Sulfating Lime N-1359
For Use in Fixed Bed Reactor Studies
Lime N-1359 was used in fixed bed reactor studies to investigate
,
the NO-S02-CaO reaction system.
This lime was prepared in the Esso FBC
by batch-charging the desired amount of lime to the combustor and con-
tinuously feeding coal.
The partially-sulfated lime was removed from
the combustor at the sulfation level at which it was still capable of
removing 90 percent of the S02 from the flue gas.
made to prepare this stone are given in the following table.
Details of the runs
Two
batches of stone were prepared and used.
Run Number 6 17
Run Length, h rs. 4.0 3.9
Coal Feed Rate, lbs/hr. 0.90 0.93
Settled Bed Height, Inches 6.0 6.0
Superficial Gas Velocity, fps 3.0 3.0
Percentage Stoichiometric Air 114. 110.
Flue Gas Composition (dry)
05' % 4.5 4.0
C , ppm 350 370
Characteristics of Lime
Removed from Combustor
CaO Utilization, % 16.6 16.8
S_O~ Removal, % 90. 90.
( dp , f 930 930
-------
Unclassified
~SS~
Security Cla..ification -1'h i!:: P a I!e
DOCUMENT CONTROL DATA. R & D
(Security clo..,f/collon 01 IIt'e, body 01 ob./roc' ond Indedn, _no 10 "'" .uel H 0,,/8red wtt- lit. o..ro" ,..." ,. cto.."'.dJ
I. OlUGINA TlNG AC TI YI TY (Corpo,.,. ou/hor) 28. ".~O"T 8.CU"ITY CLA881~ICATION
Esso Research and Engineering Company Unclassified
P.O. Box 8 2b. G"OU~
Linden New Jersev 07036 N/A
3. "EPO"T TITLE
Studies of the F1 uidized Lime - Bed Coal Combustion Desulfurization System
4. O..C"IPTIVII NOTII. (7')-". 01 reporl _d '"c'u.'". det..)
Final Report - January 1. 1971 - December 31. 1971
8. AU THO"'" (,Ir.t --. ..,ddl. ',,'tlo'. 'o.t "0..)
A1 vin Skopp Gene A. Hammons
Melvyn S. Nutkis Rene R. Bertrand
e. "ItPO"T DATIt 78. TOTAL NO. O~ PAG.. rb. ND. 7~ ".~8
December 1971 118
... CONT"ACT 0" G"ANT NO. ...O"IGINATO"'. ".PO"T NU"'."C8'
CPA 70-19 GRU.13GFS. 71
b. P"O.JIICT NO.
c. 8b. OTH." "II~O"T NOC.' (Anp olfler --,. ..., 88J' be ...,,,od
tltl. ..""rt)
d.
'o. OIlT"..UTION 8TATIIMENT
II. IUPPLIIMENTA"Y NOTEI 12. Sponsoring ActlVl ty
Office of Air Programs
Environmental Protection Agency
13. A.8T"ACT
The preliminary design of 650 KW pressurized (10 atmosphere) fluid
bed coal combustor is described. The system will consist of a 12-inch ID
combustor and a 5-inch ID regenerator, with provisions fo r con tin uous solids
circulation between the two reactors.
The results of an experimental study conducted with the objective
of determining methods of simultaneously obtaining low SOx and NOx emissions
from a fluidized bed coal combustor are reported. In fixed bed reactor
studies, a reaction system involving NO, S02 and CaO was identified. The
apparent order of this reaction was about 0.5 and it had a negative temper-
at ure dependence. The reduction of NO by CO over CaS04 also studied in
fixed bed experiments and found to proceed at app re ciab1e rates in the
13000F to 17000F, with an apparent activation energy of 6 to 8 kca1/pole.
ESSO
1473
Unclassified
IItNntr CI8.8III-tIiIi - This Page
-------
Unclassified
I '
I
Security Cla..UicaUon - This Page
, ... LIMK ... LINK. LINK C
Kav WO"DI
"OLa WT "OL. WT "OLa WT
Desulfurization
Fluidization
Air Pollution
Pollution
Sulfur Oxides
Nitrogen Oxide s
Coal
Combustion
Un.classified
fteMmty ("1...m,..Unn - Th i sPa e;e
------- |