United States
Environmental Protection
Agency
Air and Energy Engineering
Research Laboratory
Research Triangle Park NC 27711
Research and Development
EPA/600/S7-85/027 Aug. 1985
&ERA Project Summary
Fundamental Studies of
Sorbent Calcination and
Sulfation for 862 Control from
Coal-Fired Boilers
J. A. Cole, W. D. Clark, M. P. Heap, J. C. Kramlich, G. S. Samuelsen, and
W. R. Seeker
Results are presented from a labora-
tory-scale investigation of the reactivity
of calcium-based sorbents for SO2 cap-
ture after calcination at furnace opera-
ting temperatures (1200-1950°C). This
work was undertaken to provide fun-
damental information for developing
SO2 emission control technology in
pulverized-coal-fired utility boilers.
Pulverized sorbents «100 Aim diameter)
were calcined by injection into a labor-
atory gas flame reactor. Experimental
variables were time, temperature, gas
composition, limestone type, and par-
ticle size. Samples were collected for
analysis of surface area, extent of
calcination, particle size distribution,
and reactivity toward SO2.
Also investigated were fuel-rich sulfur
capture, regeneration of sulfur species
into the gas phase, and the effects of fly
ash on specific surface area and SO2
capture.
Particle heating, calcination, and sur-
face area development occurred typi-
cally in 25-35 ms. Measured surface
areas increased with decreasing calci-
nation temperature; the range for a
calcite, Vicron 45-3, was 3-15 mVg at
1200-1830°C. Surface areas for dolo-
mite reached 25 mz/g. The general
order of SO2 reactivity was dolomite
>calcium hydroxide> calcite. Fly ash
materials reduced both the surface area
and the SO2 reactivity of the sorbents
tested: calcite was affected the most,
and dolomite the least. An approximate-
ly linear correlation was found between
SO2 reactivity and specific surface area
which covered both limestones and
limestone/mineral mixtures.
This Project Summary was developed
by EPA's Air and Energy Engineering
Research Laboratory, Research Triangle
Park, NC, to announce key findings of
the research project that is fully docu-
mented in a separate report of the same
title (see Project Report ordering infor-
mation at back).
Introduction
Injecting pulverized limestone sorbents
into the radiant section of a pulverized-
coal-fired boiler, as a way to control SO2,
is being investigated by EPA. This program
is largely a development effort designed
to elucidate the principal controlling
parameters of S02 capture and, as a
result, allow for process optimization and
generalization.
The goal of the present task is to obtain
information on the high-temperature
short-residence-time reactions involving
sorbents under conditions typical of those
existing in coal flames.
The objectives addressed are:
1. To determine the physical and
thermal phenomena occurring dur-
ing high-temperature calcination of
calcium-based sorbents.
2. To determine the interrelationship
between these physical changes
and the ability of the sorbent to
absorb sulfur compounds.
3. To determine the impact of con-
trollable parameters on physical
structure changes; e.g., sorbent
type, sorbent size, temperature,
stoichiometry, and gas-phase com-
position.
-------
The approach employed in this effort was
to inject a variety of pulverized sorbents
into the high-temperature region of a
laboratory gas flame. By providing high
temperatures (>1200°C) and moderate
concentrations of H20 and CO2, the gas
flame simulated the environment in the
radiant zone of a boiler furnace. Several
diagnostic techniques were used to relate
the changes occurring during calcination
to the ability of the sorbents to absorb
SO2. Both in-situ and laboratory analyses
were performed in determining the phys-
ical and chemical changes of the particles
during the short residence times at high
temperatures. The experimental condi-
tions, controlled by the laboratory flame
environment, were varied to assess the
impact of gas-phase temperature and
stoichiometry on sorbent properties.
Measured physical changes were com-
pared with the ability of the sorbent to
absorb S02. In this manner, physical
changes were linked to the sulfur reactiv-
ity of the sorbent.
Experimental Apparatus and
Techniques
The experimental phase of this program
involved injecting calcium-based sorbents
into high-temperature flames. Physical
and chemical processes occurring as a
result of sorbent injection into a flame
were studied on-line and by subsequent
laboratory analysis. Temperature and
time in the flame were the most important
factors in determining the eventual char-
acteristics of the sorbent.
Sorbent particles were injected into a
one-dimensional laminar flame reactor.
This flame thermal decomposition reactor
(TDR) consists of a sintered bronze, flat-
flame burner downfired into a 10-cm
square stainless steel chimney. High
quality fused silica windows are mounted
on two opposing sides of the chimney for
visualization measurements. The chim-
ney also provides access for thermocouple
measurements as well as sampling, both
through ports in the wall and through the
bottom of the chimney.
Batch sampling of solids from the flame
reactor for physical and chemical anal-
yses was performed with an isokinetic
water-cooled stainless-steel probe. The
solids were collected in a large-volume
filter holder using filter elements with 0.8
um pore size. The probe was operated
above the dew point of the sampled gas
(~70°C) by restricting the cooling water
flow rate. The filter was maintained at
100°C by keeping it in an electric oven.
To determine the ability of flame-
injected sorbents to uptake S02, a
2
"dispersed-phase S02 reactivity probe"
was constructed. This probe extracted
samples from the TOR in the dispersed
phase and quenched them to 650°C to
prevent further calcination or sintering.
The sorbent stream was then drawn into
a heated zone where it was mixed with
SO2 and allowed to react. After a fixed
residence time, the sorbent passed
through a cooled zone and was collected
on a microporous glass fiber filter. The
reaction zone was heated by a tube
furnace which maintained a temperature
of 1100 ± 30°C at its midpoint. These
conditions were reproducible independ-
ent of the TDR flame condition.
Limestone sorbents were generally
selected from those used previously in
pilot- or bench-scale sorbent study pro-
grams. A high-calcium limestone, Vicron
45-3 (mean size 11 fjm), was used as the
base-case sorbent for this program. Also
employed were calcium hydroxide (mean
size 12.5 Aim), a fertilizer grade dolomite
(34 um), Marblewhite 125 (30 pm,
essentially a larger size cut of Vicron), and
a Michigan marl (18 /jm), an impure
limestone containing metal oxides and an
organic component (approximately CH0.s
N0.o4i) which accounts for nearly 5 percent
of the sorbent mass.
Results and Discussion
This section is divided into: calcination
studies, surface area studies, sulfur cap-
ture, alternate sorbents, mineral matter
effects, and a summary.
Calcination Studies
Prior to this study there was a distinct
absence of time-to-calcine data for pul-
verized sorbents for temperatures and
gas-phase compositions similar to those
found in the radiant zone of pulverized-
coal-fired furnaces. Much of the previous
data was collected at lower temperatures
and/or with larger particle sizes in inert
(typically N2 or air) atmospheres. To
investigate the times to calcine under
conditions representative of those in
furnaces, Vicron 45-3 limestone was
injected into a series of hydrocarbon and
hydrogen flames in the TDR. Solid
samples were then collected at various
distances (residence times) and analyzed
for degree of calcination. Figure 1 shows
the calcination of Vicron 45-3 at various
residence times in four methane/air
flames. The rate of calcination increases
with increasing peak temperatures, and
significant extents of calcination are
attained in times less than 100 ms at all
temperatures.
WO -
'T= 1830°C J
(3350¥)T=1630°
'<3000°F>
= 1360°C
<2500°F) ~
Vicron 45-3
1830°C
O 1630°C
A 1SJS°C
O 1360°C
50 100 150
Residence Time, ms
Figure J. Calcination dependency on
residency time for Vicron 45-3
injected into four methane/air
flames in the TDR.
The present data are consistent with
those of earlier studies. Figure 2 shows
an Arrhenius plot of the present data as
well as earlier differential reactor data
and dispersed-phase data. The zero-order
kinetic rate constant recommended earl-
ier predicts a shorter time than was
observed for calcination at furnace tem-
peratures because, at these tempera-
tures, the particle heatup time has be-
come a significant portion of the total
time required for calcination.
CO
c'
O
;cr
•5 10°
"<5
O 70'
8 102
» W3
• This Study Vicron 45-3
Dispersed-Phase
Differential
Reactor Data
Limestone
Furnace
. Range
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2
tow, AT'
Figure 2. Comparison of present study
with other data for calcination of
Vicron 45-3 and similar lime-
stones.
Surface Area Studies
Several sorbent properties are believed
to have significant influence on sorbent A
reactivity toward S02, including pore size, *
total porosity, total specific surface area.
-------
anddegreeof crystallinity. Also important
is the initial sorbent particle size, which is
chosen at the outset. These properties,
however, manifest themselves in the total
specific surface area, which is a readily
measurable parameter.
Samples for surf ace area analyses were
collected at various residence times for
Vicron 45-3 injected into four methane/
air flames identical to those represented
in Figure 1. The data are shown in Figure
3. The surface areas measured here (5-
10 mVg) are low relative to those reported
for similar stones in lower temperature
work and suggest that the calcine is not
very reactive.
To test fragmentation of the sorbent
during calcination, three samples of
Vicron 45-3 were collected in the TOR:
the first, collected without a flame; and
the other two, collected at long residence
time (>150 ms) from methane flames
providing peak particle temperatures of
1360°C and 1830°C. All three sorbent
size distributions, as determined by X-ray
sedimentation, are shown in Figure 4.
This shows that fragmentation does
occur, and that it increases with increas-
ing temperature. This degree of fragmen-
tation is inadequate to account for the
increase in surface area during calcina-
tion.
Sulfur Capture
If limestone activation is related to
surface area. Figure 3 suggests that
reactivity (or calcium utilization) should
increase as the peak calcination temper-
ature is reduced. Sulfation test results for
Vicron 45-3 in the dispersed-phase SO2
reactivity probe, immediately following
calcination in the TDR, are shown in
Figure 5. Although the utilization effi-
ciency does increase somewhat with
decreasing calcination temperature, the
effect is not as dramatic as the differences
in surface areas might suggest. The
scatter in the low-temperature data
occurs because the ability of the sorbent
to calcine effectively is more sensitive to
spurious pulses in the sorbent feed rate at
these conditions. The higher utilization
values are considered more reliable be-
cause those samples showed extents of
calcination more representative of pre-
vious data for samples collected under
identical conditions.
Experiments were conducted, with the
TDR under both reducing and oxidizing
conditions, to determine if sulfur capture
at a given flame temperature depended
an its capture as HjS (to form CaS, as in
rthe reducing region of a burner flame) or
as S02(to form CaSCU, as in the oxidizing
region of a burner environment). In these
tests, the TDR was operated with a H2S-
doped methane flame, with flame stoi-
chiometry varied (from oxidizing to re-
ducing) by substituting nitrogen for some
of the combustion air. The results—over
the temperature range 1250 to 1400°C—
show that, at a given temperature, cal-
cium utilization after 150 ms residence
time is essentially independent of sulfur
captured as H2S or SO2.
Although sulfur capture by limestone
under fuel-rich conditions occurs with
effectiveness similar to fuel-lean capture.
one potential limitation is the extent to
which the fuel-rich product, CaS, might
be oxidized to CaO and S02 in the
subsequent fuel-lean region of an actual
staged combustion boiler furnace. To test
for this, a powdered CaS, similar in size to
Vicron 45-3, was injected into four fuel-
lean methane/air flames. Insufficient
data were collected for a detailed analysis
of the kinetics of the oxidation reaction,
but the data in Figure 6 demonstrate the
rapid regeneration of most of the sulfide
at furnace temperatures. This level of
sulfide regeneration in such a short time
70.0 -
.o
Tmax=1630°C
Tmax = 1830°C
SO 100
Estimated Residence Time, ms
150
Figure 3. Specific surface area for Vicron 45-3 injected into methane/air flames.
•S
a
I
100
90
80
70
60
SO
40
30
20
10
\ I '
Vicron 45-3
1830°C
1360° C Peak
Flame Temperature
No
Flame
Figure 4.
50 20 10 5 21
Equivalent Spherical Diameter, (im
Effect of peak calcination temperature on the ultimate particle size distribution of
Vicron 45-3 injected into methane/air flames.
-------
15
I
"
1200 1300 1400 1500
Peak Calcination Temperature in TDR, °C
1600
Figure S.
1OO
Calcium utilization efficiency of Vicron 45-3 in reactivity probe. Residence time, 600
ms; nominal reaction zone temperature. 1100°C; 6% SO2 by volume.
100
Residence Time, ms
150
Figure 6. Regeneration of CaS with time in the TDR for fuel-lean methane flames.
(<100 ms) may render the fuel-rich
capture process unacceptable unless
solids are removed before tertiary air
injection.
Alternate S or bents
The calcitic limestone (Vicron 45-3)
was used as a baseline because of its
high calcium content, low cost, and the
general availability of similar stones.
4
Sorbents other than Vicron 45-3 may
respond quite differently when injected
into high-temperature gases. In addition,
their reactivity toward S02 sorption may
relate to specific surface area in a differ-
ent way. The sorbents chosen for study
here have been used previously in bench-
and pilot-scale studies and are represen-
tative of the major classes of limestones.
In addition to Vicron 45-3, another high-
purity calcite having a larger particle size,
Marblewhite 125, was investigated.
Other calcium-based stones (marl, dolo-
mite, and calcium hydroxide) were tested
also.
Figure 7 shows the trends in surface
area development and calcination for four
alternate sorbents in a 1515°C methane/
air flame in the TDR. In this flame the
residence times were longer enough in
some cases that a slight decline in specific
surface area is suggested for all of the
sorbents except dolomite. In addition, the
extent to calcination rises rapidly at short
times for all sorbents, corresponding to
the initial particle temperature rise time
of about 80 ms. After the initial rise the
extent of calcination increased only slowly
in the relatively cooler downstream gases.
The effect of temperature on surface area
was qualitatively the same for the alter-
nate sorbents as for Vicron 45-3, increas-
ing with decreasing flame temperature.
Two of the alternate sorbents, dolomite
and calcium hydroxide, were also tested
for sulfation in thedispersed-phase reac-
tivity probe following calcination in the
TDR. Figure 8 compares the results of
these tests with Vicron 45-3 data and
with the (unsulfated) surface areas
measured when the sorbents were cal-
cined in theTDR under similar conditions
of TDR residence time and temperature.
For each sorbent, both specific surface
area and calcium utilization efficiency
increased monotonically with decreasing
calcination temperature. However, al-
though Ca(0 H)2 and Vicron 45-3 had very
similar surface areas over the range of
study, Ca(OH)2 had a significantly higher
calcium utilization efficiency.
Mineral Matter Effects
The effects of coal-ash on limestone
reactivity were investigated. To simulate
the effects of ash, kaolin, a mineral
common in coal ash, was added to Vicron
45-3 to make a 5 percent by weight
mixture. The mixture was then injected
into several flames and samples were
collected for surface area analysis. The
results are shown in Figure 9. Kaolin
severely reduced the specific surf ace area
of Vicron 45-3 at every condition studied.
Several sorbent mineral mixtures were
also tested for sulfation in the dispersed-
phase SOz reactivity probe. Figure 10
shows the effect of kaolin on the reactiv-
ities of dolomite, Ca(OH)2, and Vicron 45-
3. These data should be compared with
the results for the pure sorbents in Figure
8. Kaolin has essentially no effect on
dolomite and Ca(OH)2. However, the
utilization of Vicron 45-3 was severely
-------
tO
50 100 150
Resilience Time, ms
50 100 150
Residence Time, ms
200
Figure 7.
Specific surface area (A) and percent calcination of alternate sorbents (B). Peak
calcination temperature 1515°C in the TOR.
reduced, \n parallel with the effect that
kaolin had on surface area.
Summary
Measurements of the specific surface
area of heat-treated sorbents, and of the
reactivity of such sorbents for capturing
sulfur, have indicated the following
points: 1. The physical structure of sor-
bents (particle size, extent of calcination,
specific surface area) after high-tempera-
ture heat treatment depends on the
sorbent type and peak calcination tem-
perature. 2. The specific surface area as
measured by BET techniques provides a
reasonable indication of the ability of the
calcined sorbent to capture sulfur. 3. The
specific surface areas measured for dif-
ferent sorbents calcined under furnace
conditions (1200-1800°C) are relatively
low (4-15 mVg) and are inversely related
to the maximum temperature that the
sorbent has experienced.
The overall effect of surface area on
sorbent reactivity toward S02 is shown in
Figure 11. Although there may be indi-
vidual trends for each sorbent and
sorbent/mineral mixture, the data as a
whole demonstrate that reactivity is
directly a function of specific surface area
after calcination for all sorbents studied.
Conclusions
In this study, the high-temperature
short-time behaviors of calcium-contain-
ing sorbents were investigated in an
environment which simulated the radiant
zone of a pulverized-coal furnace. The
1000 1200 1400 1600 1800
Temperature, °C
20
Calcium
' Hydroxide
Dolomite
Vicron 45-3
1000 1200 1400 1600 1800
Temperature. °C
Figure 8. Comparison of surface area
development and calcium util-
ization efficiency as a function
of calcination temperature for
three sorbents.
physical and chemical changes that influ-
ence the ability of the heat-treated
sorbents to capture sulfur were measured.
A new direct measuring technique was
developed, based on the ability of the
heat-treated sorbent to capture sulfur
without having to quench the sample to
ambient conditions. This technique util-
ized a reactivity probe which allowed the
sorbent to be subjected to S02 under
well-controlled dispersed-phase condi-
tions. A variety of calcium-containing
sorbents and mixtures of sorbent and
mineral matter were investigated.
£
D 1200°C
o 1360°C
0 1515°C
e, 1630°C
\
\
w
B a
~8
10
75 20 25 30
Wt-Percent Kaolinite in Vicron 45-3 Before
Calcination
Figure 9. Effect of temperature on the
ultimate specific surface areas
of calcined Vicron 45-3/kaolin
mixtures.
50
8 40
§ 30
.u
I
•g 20
%
10
,
-------
I
40
30
20
10
I I
Reactivity Probe Conditions:
Residence Time - 600 ms
T=1JOO°C
6% S02
Vicron 45-3
Dolomite
CafOHh
5% SiOx in Vicron
5% Kaolin in Vicron
10 15 20
Specific Surface Area, rrf/g
25
Figure 11.
Calcium utilization efficiency of sorbents and sorbent/'mineral mixtures as a
function of specific surface area of sample calcined under identical conditions.
Increasing specific surface area corresponds to decreasing peak calcination
temperature for a given sorbent.
Conclusions from the results of this
study are:
• At furnace temperatures (1200-
1800°C) pulverized sorbents heat and
calcine rapidly (<100 ms), although
larger particles (>50 /am) may exper-
ience longer calcination times.
• Surface area develops in parallel
with—and as a result of—calcination.
For a given sorbent, higher surface
areas are achieved at lower calcination
temperatures.
• Thermal comminution of limestone is
evident during calcination. However, it
alone is not sufficient to account for
the measured increase in specific
surface area.
• Reactivity for sulfation, for a given
sorbent, depends on the specific sur-
face area of the sorbent after calci-
nation, prior to exposure to S02.
• Certain minerals will effect a strong
decrease in both surface area and SO2
reactivity after calcination. The result
is sorbent dependent.
•fr U. S. GOVERNMENT PRINTING OfFtCt 1985/539-111/20639
-------
J. A. Cole, W. D. Clark, M. P. Heap, J. C. Kramlich. G. S. Samuelsen. and W. P.
Seeker are with Energy and Environmental Research Corp., Irvine, CA
92714-4190.
G. Blair Martin is the EPA Project Officer (see below).
The complete report, entitled "Fundamental Studies ofSorbent Calcination and
Sutfation for S02 Control from Coal-Fired Boilers," (Order No. PB 85-221
729/AS; Cost: $17.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Air and Energy Engineering Research Laboratory
U.S. Environmental Protection Agency
Research Triangle Park, NC 27711
United States Center for Environmental Research
Environmental Protection Information
Agency Cincinnati OH 45268
Official Business
Penalty for Private Use $300
EPA/600/S7-85/027
OC00329 PS
U S ENVIR PROTECTION AGENCY
REGION 5 LIBRARY
230 S DEARBORN STREET
CHICAGO IL 60C04
------- |