SWELLING AND AGGLOMERATION EFFECTS FOR BITUMINOUS COAL
IN A LAMINAR FLOW-REACTOR
by
John Francis Dolan
Submitted in partial fulfillment
of the requirements for the
degree of
Bachelor of Science
at the
Massachusetts Institute of Technology
May,
1980
)John Francis Dolan May, 1980
The author hereby grants to M.I.T. permission to reproduce and
to distribute copies of this thesis document in whole or in part.
Signature of
DepartmenW of Chemical Engineering, May 8,1980
Certified by
- -
Chairman,
Accepted by.......
'00ý01ý
--'
..
.
. . .
. .
..
a.
.a &
...*
Thesis Supervisor
&
Chairman, Department Committee
ARCHPVES
NOV 01 1994
ABSTRACT
SWELLING AND AGGLOMERATION EFFECTS FOR BITUMINOUS COAL
IN A LAMINAR FLOW REACTOR
by John F. Dolan
Submitted to the Department of Chemical Engineering
in May, 1980, in the partial fulfillment of the
requirement for the degree of
Bachelor of Science
Swelling and agglomeration effects in bituminous
coal particles under pyrolytic and hydropyrolytic conditions were studied.
Experiments were carried out on.
coal particles of mean diameter 22.5 and 49,um at 600 and
10000 C at various feed rates.
Particle counts were per-
formed to determine agglomeration effects.
The extent of
particle swelling was assessed by particle sizing.
fective particle diameters were calculated.
was found to be significant for 20-25 .ff
pyrolytic
Swelling
particles under
conditions at 6000 C and hydropyrolytic
ditions at 10000 C.
Ef-
con-
Agglomeration effects were significant
in all cases tested.
Thesis Supervisor:
Jack B. Howard
Professor of Chemical Engineering
TABLE OF CONTENTS
Section
1.0
2.0
3.0
3.1
4.0
4.1
5.0
Page
Introduction
General Apparatus and Run Procedure
Procedure for Agglomeration Analysis
Results and Discussion on Agglomeration
Procedure for Swelling Analysis
Results and Discussion on Swelling
5
6
7
8
11
11
Overall Effect on Particle Size:
Results and Discussion
13
6.0
Conclusions and Recommendations
Appendices
A
B
C
Data Tables
Table 1: Agglomeration vs. Feed Rate
for 20-25 mm particles, 600OC,He
Table 2: Agglomeration vs. Feed Rate
for 20-25pm particles, 10000C, He
Table 3: Agglomeration vs. Feed: Rate
For 20-25 tm particles, 1000oC, H
2
Table 4: Agglomeration vs. Feeds:Rate
for 45-53 .,m particles, 6000C, He
Table 5: Agglomeration vs. Feed Rate
for 45-53 ,um particles, 100000, He
Table 6: Agglomeration vs. Feed Rate
for 45-53 km particles, 1000 0 c, H2
Table 7: Swelling Factor, Weight Fraction
Recovery, and Particle Size vs. Run
Condition
Table 8: Particle Mean Diameter and
Standard Deviation vs. Feed Rate
Table 9: Data for Agglomeration
Experiments
Table 10 : Particle Size Data
Error Analysis for Particle Agglomeration
and Swelling Experiments
Figures
-
Bibliography and Footnotes
14
15
15
15
15
15
15
15
16
17
18
21
23
24
LIST OF FIGURES
Page
Number
1
Detailed Schematic of the Reactor
and Peripheries
24
2
Agglomeration vs. Feed Rate for
25
20-25 um particles 600oC, He
3
Agglomeration vs. Feed Rate for
20-254m particles 1000 0 C, He
26
4
Agglomeration vs. Feed Rate for
45-53mm particles
6000C, He
Agglomeration vs. Feed Rate for
4 5-53Am particles 1000oC , He
Agglomeration vs. Feed Rate for
45-53 Am particles 10000C , H 2
Linearized Plots of Agglomeration
vs. Feed Rate
Micrograph of 20-25pm Feed
Micrograph of 20-25 ,m ,
6000 C He
27
5
6
7
8
9
28
29
30
31
31
10
Micrograph of 20-25 Am,
10000 C , He
32
11
Micrograph of 20-25,Lm,
10000 C , H2
Micrograph of 45-53Am Feed
32
12
m ,
13
Micrograph of 45-53
6000 C He
14
Micr 8 graph of 45-53 4m,
1000 C He
Micrograph of 45-53 }m
10000 C H2
15
33
33
34
34
1.0
Introduction
Gas yields have been found to depend on particle
size in the pyrolysis and hydropyrolysis of bituminous
coal.
For this reason , determination of the true particle
diameter during the pyrolysis process is needed to model
the reaction kinetics.
We know that particle size is influenced by two
effects: swelling and agglomeration.
Experiments have been
conducted to determine the extent of swelling in coal
pyrolysis.
Pohl et. al. 1 have found extensive swelling
for bituminous coal particles at temperatures below about
13000 C.
Sung 2 has found agglomeration of coal particles
to be quite important as well.
The main objective of this research was to determine
the effective particle diameter in a laminar flow coal
gasification reactor under various conditions and to use
these data to predict diameters at Other conditions.
The
effective particle diameter, if different from the original
feed diameter,is determined by two phenomena: particle
swelling and particle fusion. The effects of the following
four parameters on the extent of agglomeration and swelling
were determined: temperature, type of main gas (He or H 2 )
coal feed rate, and particle size.
2.0
General Apparatus and Run Procedure
The apparatus used in these experiments consisted
of a Vertical tube furnace in which a steady , laminar flow
of preheated gas was maintained.(see Fig. 1) A feed sample
was metered by a powder feeding device into
a water-cooled
tube which extended through the furnace shell along the
centerline of the reaction tube to the furnace hot zone.
The particle stream emerging from the water-cooled feed tube
was heated rapidly to the reaction temperature and entrained
along the furnace centerline for a desired distance before
it was directed through a heated probe and collected in
sample dishes placed at the furnace exit.
3.0 Procedure for Agglomeration Analysis
Extents of agglomeration were assessed by measuring
the difference in the number of particles per milligram
before and after a run.
A microscope and balance were used.
An estimated one milligram of sample was transferred
with a microspatula onto a square cut from graph paper.
The paper was folded on all sides to keep the sample from
sliding off the enclosed grid.
The grid consisted of 900
(30 X 30) squareseach imm2 in area.
The sample was spread
out as evenly as possible on the grid using a small brush.
At this point the sample was ready for counting
under a microscope.
A 30X Wetzlar microscope equipped with
a high intensity lamp was used.
The number of particles in
twenty random sets of two adjacent squares were counted and
the average of the counts was multiplied by 450 (as there
are 450 sets of two squares on the grid).
The grid was then
placed on a Mettler H20 balance and weighed to an accuracy
of 0.01 mg.
The sample was then brushed off the grid so
that only the paper remained.
again.
The grid was then weighed
The total number of particles divided by the dif-
ference between these two weight measurements yielded the
number of particles per mg. for a given sample.
3.1 Results and Discussion
Experiments were run under three general conditions
for each of the two particle sizes: helium main gas at
600 and 10000 C and hydrogen main gas at 10000 C. Table 9
contains all particle count data.
Data related to weight
loss upon pyrolysis and hydropyrolysis is presented in
Table 7 .
Tables 1-6 show the calculated agglomeration
factors for each condition where
Agglomeration
Factor
(Initial SamQle Wt.)
/Feed Particles/mg)
Product Particles/mg) (Final Sample Wt.)
Agglomeration has the effect of increasing the average particle
diameter as it decreases the number of particles.
An ag-
glomeration factor represents the effect that agglomeration
has on the average particle diameter during devolatilization.
For example, an agglomeration factor of 1.5 means that agglomeration effects have increased the average particle diameter
by 50%.
The agglomeration factors listed in Tables 1-6 show
significant agglomeration effects for every general case.
Although not true in all six cases, a definite proportionality exists between the agglomeration factor and the feed
rate for the 20-25 }m He runs at 10000 C as well as for the
45-53 mm runs at both 600 and 10000 C.
This proportionality
can clearly be seen in Figures 3,4, and 5 which plot the agglomeration factor as a function of feed rate for the above
three conditions.
Figures 2 and 6 show that for the 6000 C, He,
20-25 pm and 10000 C, H2 , 45-53 Pm particle runs, no functionality exists between the agglomeration factor and the
feed rate.
These results indicate the possibility of an
agglomeration maximum for these conditions.
FigUre 7 contains
"best fit" lines for the agglom-
eration vs. feed rate data presented in Tables 1-6.
These
lines show that in both the case of 20-25 um and 14.553.tm particles, agglomeration effects are greater at the
higher temperature (10000 C) over the range of feed rates
tested.
Due to the variability in the calculated agglomeration fa'ctors for the 10000 C , H2 , 20-25Am particle
runs, the effect of substitution of hydrogen for helium
as the carrier gas is unclear for this particle size.
For
the 45-53 gm particles, extents of agglomeration were
lowered substantially by this substitution.
Larger temp-
erature and concentration gradients are present in the
4 5-53,om
particles than in the 20-25km size.
This effect,
combined with the highly reactive H 2 gas., created conditions
that tended to reduce the particle stickiness.
Figure 7 shows that for a given temperature and carrier
gas
, the agglomeration effect is greater for the 20-
25 ~m than the 45-53 pm particles at all feed rates tested.
10
The extent of agglomeration will be proportional to
the number of collisions multiplied by the fraction of
collisions resulting in agglomeration.
Since for a given
particle size , feed rate , and gas flow rate, the number
of collisions should be the same regardless of the temperature or carrier gas, the great variability in the "best
fit" agglomeration lines in Figure 7 indicate that the
fraction of collisions resulting in agglomeration (i.e.
the relative particle stickiness) can vary substantially
for different conditions.
4.0
Procedure for Swelling Analysis
A microscope and camera were used to assess the extent
of swelling.
A sample was transferred with a microspatula
onto a slide and spread evenly using a small brush.
A
Microstar light microscope equipped with a camera capable
of taking Polaroid Type 55 pictures was used to photograph
the sample slides.
The 6000 C helium
4 5-53sm
run samples
were photographed at a magnification of approximately ~40X
while all others were photographed at approximately 100X.
Two good pictures of each sample were taken.
A ruler was
used to measure particle diameters off the photographs.
Measurements were made accurately enough to determine the
size category into which a given particle should be catalogued.
Comparision of this average with that of the feed
sample yielded the swelling factor.
4.1 Results and Discussion
Experiments were run under three general conditions
for each of the two particle sizes: helium runs at 600 and
10000 C and a hydrogen run at 10000 C .
Table
10
contains
particle sizing data used to calculate the swelling factor
for each condition where
product diameter
Average feed diameter
1
SAverage
Swelling factor =X
agglomeration
factor
A swelling factor represents the effect that swelling has
on the average diameter during devolatilization.
For
example , a swelling factor of 1.5 means that swelling
About 60 particles per photograph were measured.
effects have increased the average diameter by 50%.
Table 7
lists the swelling factor calculated for each of the above
six cases.
Swelling decreased with increased temperature for
both particle sizes.
This trend is verified by Pohl et.
al.1 who found no particle swelling above 13000 C. The use
of H2 instead of He as the carrier gas increased the swelling
factor in both the 20-25 Am and the 45-53,xm particles.
For
the smaller particles the increase was quite substantial;
from 0.98 to 1.16.
Table 7 also shows the swelling factors for the 2025,4m particles to be substantially greater than those for
the 45-53/Am particles.
In fact, the 45-53 Pm particle
runs at 10000 C with either carrier gas showed particle
shrinkage.
This is due to the high weight losses (ap-
proximately 6%
experienced at this temperature.
5.0
Overall Effect on Particle Size
Results and Discussion
Table 7 lists the mean and standard deviation of
particle diameters calculated from the particle size
distribution data listed in Table 10 .
Photographs
of feed and product particles are presented in Figures
8-15.
Due to the irregularity of particle shapes, it
was difficult to assess the true average diameter in the
20-25,~m or
4 5-53AMm
feed.
These averages were taken to
be 22.5,um and 49Am respectively and were then used as
a basis to size product particles.
Table 7 shows the mean product diameters for
the 20-25.m feed particles to be almost doubled upon devolatization.
The average diameter for the 10000 C , H2
runs increased by over 250% to 56.9pm.
Devolatilization
effects on the 45-53,pm feed mean particle diameters were
not as substantial.
In fact, for the 10000 C , He runs,
the mean diameter was essentially unchanged.
The mean particle diameter and standard deviation
were calculated for each feed rate for both the 20-25,Am
and the 45-53gm particles at 6000 C with He main gas.
These calculations are presented in Table 8.
Standard
deviations generally rose with increased agglomeration
and swelling as expected.
14
6.0
Conclusions and Recommendations
For 20-25um particles, agglomeration effects are
significant at every condition tested (feed rates of
11-29 mg/min, He or H 2 as an entraining gas, 600 or
10000 C).
Swelling effects were important under pyroly-
tic conditions at 6000 C and hydropyrolytic
conditions
at 10000 C.
For
4 5-53,"m
particles, agglomeration effects are
significant at all conditions tested.
However, swelling
is negligible under all conditions tested.
At 10000 C,
with either carrier gas, these particles actually decreased
in size as a result of high weight loss..
At temperatures greater than 6000 C , swelling
decreases with increasing temperature.
In general, 20-25um particles are more susceptible
than 45-53 pm particles to swelling and agglomeration effects.
The highly reactive H 2 environment increases swelling
and agglomeration effects for 20 -25
4
m particles at 10000 C.
In order to obtain a sample as free from agglomeration
and swelling as possible , a 45-53 ,um particle run at 10000 C
under pyrolysis conditions at a low feed rate should be
performed.
Data from pyrolysis and hydropyrolysis experiments
run at 8000 C would provide information for a deeper analysis of temperature effects on swelling and agglomeration.
ADpendix A:
Data Tables
Table 1
20-25 Am 6000 C He
Agglomeration
Feed Rate
Factor
(mg/min)
1.44
11.9
1.44
1.40
1.40
13.5
19.7
23.8
Table 3
20-25 um 10000 C H,
Agglomeration
Feed Rate
(mg/min)
Factor
r
15.0
2.36
28.4
1.98
Table 2
20-25 um 10000 C He
Agglomeration
Feed Rate
(mg/min)
Factor
15.0
1.75
17.2
1.91
21.3
1.92
28.4
2.01
Table 4
45-53,m 6000 C He
Feed Rate
Agglomeration
(mg/min)
Factor
6.9
1.11
1.24
15.3
26.9
31.7
1.26
1.26
41.3
1.36
Table 5
45-53um
Feed Rate
(mg/min)
Table 6
1000 0
C He
21.1
Agglomeration
Factor
1.28
1.48
29.3
1.57
32.5
1.40
15.9
4 5-53
m• 10000 C H9
Feed Rate
(mg/min)
15.9
26.9
32.5
Agglomeration
.Factor
1.20
1.14
1.15
Table 7 :Swelling Factor, Weight Fraction Recovery,
and Particle Size vs. Run Condition
T
Final
Sample wt.
Sample wt.)
Initial Sample wt.
Condition (Inal
Swelling
Factor
Mean
Diameter
(0m)
20-25 m
He,600 0 C
0.61
1.25
40.3
20-25mm
o
0.47
0.98
41.7
0.42
1.16
56.9
0.63
1.01
63.7
0.44
0.718
50.7
0.40
0.768
43.1
He,1000
20-25
Am
m
He 6000 C
45-53
m
He 10000 C
45-53
Deviation
(% of mean)
41.8
C
H2 ,100 00 c
-53w
Standard
m
H2 10000 C
30.0
Table 8:
Feed Size
Particle Mean and Standard Deviation vs.
Rate at 6000 C,He Main Gas
Condition
20-25 Aum He-, 6000 C
Feed Rate
He,1000 0 C
Standard
Deviation
(% of mean)
11.9
38.4
13.5
19.7
39.1
44
42
42.0
40.8
43
38
50.3
29
61.4
30
30
35
23.8
45-53 *m
Mean Particle
Diameter
(•m)
Feed
6.9
15.3
26.9
31.7
41.3
57.8
57.8
82.1
26
Table 9
Data for agglomeration analysis
6000 C
He
Feed Rate
(mg/min)
45-53
/Am
Count
(particles/mg)
Average
(particles/mg)
7500
6500
Feed
6460
6100oo
4730
6300
6000
8170
6.9
7400
7520
6890
8270
4870
5360
15.3
6990
26.9
5260
4980
5120
31.7
5030
5210
5120
41.3
5100
4400
3070
4190
4220
He 6000 C
Feed
23000
20500
23250
11.9
13.5
19.7
10020
14280
12330
12400
13630
12860
20-25
mm
22250
12150
12370
13250
9480
23.8
15860
14260
13200
Table 9 (cont.)
He 10000 C
45-53
Feed Rate
(mg/min)
15.9
/m
Count
(particles/mg)
6840
7090
Average
(particles/mg)
6965
21.1
4430
4590
4510
29.3
3660
3800
3730
32.5
5340
5280
5310
He 10000
20-25 Am
5310
15.0
9800
8637
10800
17.2
6380
7080
6729
21.3
6580
6500
6536
28.4
5230
5671
6200
H2 10000 C 45-53 mm
8432
10320
9376
26.9
10110
12240
11175
32.5
12670
8930
10800
15.9
20
Table 9 (cont.)
20-25 tim
HII 10000 C
--
f
Feed Rate
(mg/min)
15.0
28.4
Count
(particles/mg)
Average
(particles/mg)
3830
4200
4015
7250
6230
6740
Table 10: Particle Size Data
Particle Size Distribution
6000 C He
45-53 ,m
Particle Diameter Ranges
Feed rate
(mg/min)
40-55um
0-39um
(%)
56-77um 78-99um
(W)
(%)
(%)
feed
5
68
36
2
6.9
11
55
36
12
15.3
5
38
50
46
26.9
8
36
57
30
31.7
8
48
39
30
41.3
1
25
51
62
100-121um
122-143um
(%)
(%)
45-53um
10000 C He
- Particle Diameter Ranges
Feed rate
(mg/min)
0-24um 25-34um
35-44um
(o)
(%)
15.9
14
37
32.5
8
32
45-54um
55-69um 70-88um 89-1O8um
(%)
(%)
(%)
15
30
0
1
18
39
3
0
45-53um 10000 C H2f
J
Particle Diameter Ranges
Feed rate
(mg/min)
0-24um
Wf)
25-34um
35-44um
(%)
45-54um
55-69um 70-88um
(%)
(%)
15.9
32
19
28
26.9
61
9
13
61
11
7
32.5
0
21
(%)
0
89-108um
(%)
0
22
I0
CD
N)
CD CD
-
m.,,
0 "A
o
N)
N)
00
00 ý'OA
CD
OBkM-SC
CD
NJN
H- )
'D
O'\
op
0
0.
CD
0 0
OO
SC
C
0
7\
O 0 CO - ZC
0•
,0
S--0
2
N)
0
So
N)
coP
N)
0
F"
)
N)
N)
O
CO
\0
-N)O
\z
N)
1,0 -,. 0 N U,0-•,.. H
-0
f-,
Cc+
~-~~ri
o
0
0
So
/'9
N0
,c
---
N)
P
N)
O~
h*-~*
I-~
00
C
,cD0
O\
01
/'9
cor
-0
0
-',.
CD
e
00
"0 0
Or
--,3
..- ,co--0
L.x
\0
CD
N)
C
00
POP•
t-., •
0
~0
I
ON
,----0
0
ON N
N)
*----
Co ., N) - ,O
CO
N) P
CD
CO "-0",
0
-0
2
'
0
N)
"- Cc
",
Cc
0\
P'--00
cc
ON
2
23
Appendix B
Error Analysis for Particle Agglomeration and
Swelling Experiments
Agglomeration
Errors in precision were analized.
In the agglomer-
ation counts, the largest variation for a given set of conditions was
in the case of the 20-25,~ m particles at 10000 C
using He main gas, fed at 15 mg/min (see Appendix 7.8).
In
this case ,
Variation= 0.5 Total Spread of Counts
(
Average Count
5)(490)
8637
0.318.
The variation in the 20- 2 5mm feed counts was =(0. 5 )(22250)= 0.062.
22250
Plugging the average counts into the agglomeration factor
equation yields 1.75.
Plugging in the extremes within
variations calculated above yields 2.05,
The percentage difference = 2.05-1"75
1.75
(100%)= 17.2%
Therefore 17.2%is the maximum precision error in a reported
agglomeration factor. At-least two counts were performed
in determining each agglomeration factor to lower this
error substantially.
Swelling
As a consistency test for swelling determination,the
average particle diameter in each of three sets of two
photographs of a given sample was determined.
percentage of variation was = (0.5)(61,m-
4 ,am)
The total
(100o)= 6.3%.
This percentage is small enough to provide confidence in
swelling factors calculated from the number of particles
provided by two photographs.
24
Appendix C: .Figures
Figure 1: Detailed Schematic of the. Reactor and
Peripheries
110 VAC
200A
200
five
*o
10 VAr
A
*'10
1.
20
25
i .60
1.40-
0
0
1.20
S
0
H
1.00
*
n 80
1L
I
4
I
I
I
SI1
t
1
t
I
i
10
et
i
v
t
--
7
40
Feed Rate
Figure 2
Agglomeration vs. Feed Rate
(mg/min)
20-25um 6000 C He
26
2.20 .
r
2.00
W.
-,.
9/*
/
0bfL
o
O•
0
*H
1.80 -
Cd
O
/
/
E
/
/
/
1.60+
/
/
1.40
I
I
I
r
S
!
t
i
/
)
I
~!
)
I
I
I-i
j
I
I -
I
1
- I-
I
T
20
Feed rate
(mg/min)
Figure 3
Agglomeration vs.
Feed Rate 20-25um 10000 C He
I•
I
Id
27
1.60
1.40-
D
1.20-
I
a
C
O
H
1.00
0.80
1
_1
r1
·
1
~1
I
·
I
I
·
.
.
I
I
I
I
I
-~
.~
q
1
I
__ 1
IL
I
_·
I
I
I
·
I
I
L
I
·
40
10
Feed rate
Figure 4
(mg/min)
Agglomeration vs. Feed Rate 45-53um 6000 C He
-
28
1.60 .
1.40
-
0
1.20 -
-
0
*H
4:
S
0
bL
1.00 -
.8
m
-I
O.C
I
I
·.
1
I
I
|
-
·
1 -r·
I
I
I
I·
----I
l
I
Feed rate
I
1
-----1
20
s
·
I
I
I
I
·I
I
I
·~--- ·
|
I
I
(mg/min)
Figure 5
Agglomeration vs. Feed rate 45-53um 10000 C He
-
29
1.60.
1.40.
-
-
C
0
1.20
z
0
1.20 -
S
0
H
00-
0 .80
-
-
0
_
__
__
_ _
10
Feed rate
Figure 6
I-
a
I
30
20
i
(mg/min)
Agglomeration vs. Feed rate 45-53um 10000 C H2
i
I
40
2.0
1.5
.4,.
*
* * *
-..
4
*4
4. * 0
-
.
---
-
-
_
- _*
*
-
-
1.0
20-25"m 10000 C He
-
- -
6000 C He
.
10000
C He
20-25,'m
-
45-53 m 6000 C He
0.5
,
f
I
f
I
l
I
10
1
1
1
1
1
20
I
l
I
S
Feed Rate (mg/min)
Figure 7
Linearized Plots of Agglomeration vs. Feed Rate
l
1
1
30
I 30
1
m
•
m
-IV
S
4
I
i
Figure 8
20-25/Am Feed
Figure 9
20-25Am
, 6000 C , He
Figure 10
20-25 um, 10000 C , He
fed at 21.3 mg/min
b -.
Figure 11
20-25 Mm , 10000 C , H2 fed at 28.4 mg/min
Figure 12
45-53,mm Feed
Figure 13
45-53gm ,6000 C , He fed at 26.9 mg/min
34
'·
I
--
rt:
Figure 14
45-53 um , 10000 C He fed at 32.5mg/min
ML AK
Figure 15
45-53,Mm 10000 C H 2 fed at 26.9 mg/min
35
Bibliography and Footnotes
1.
Pohl, J.H. , H. Kobayashi, and A.F. Sarofim, "The
Effects of Temperature and Time on The Swelling
of Pulverized Coal Particles", paper presented
at the Combustion Institute Technical Meeting
Boulder , Colorado
2.
(1978)
Sung, W.F., "The Study of the Swelling Property of
Bituminous Coal", S.B. Thesis , Dept. of Chemical
Engineering, Massachusetts Institute of Technology,
Cambridge, Mass., (1977)