Ind. Eng. Chem. Process Des. Dev., Vol. 17, No. 3, 1978 Rathore, R. N. S., Van Wormer, K. A., Powers, G. J., AlChE J., 20 (3), 491 (1974a). Rathore, R. N. S.,Van Wormer, K . A., Powers, G. J., AlChE J., 20 (5), 940
281
Tedder, D. W., PhD. Thesis, The University of Wisconsin-Madison, 1975. Wang, J. C., Henke, G. E., Hydrocarbon Process., 45 (e), 155 (1966).
Received for review April 29, 1977 Accepted January 26,1978 Presented at 4th International Congress in Scandinavia, Copenhagen, Denmark, Apr 18-20,1977.
(1974b). Rodrigo, F. R., Seader, J. D., AlChEJ., 21 (5), 885 (1975). Seader, J. D., Westerberg, A. W., AlChEJ., 23 (6), 951 (1977). Stupin, W. J., Lockhart. F. J., Chem. Eng. Prog., (Oct 1972).
Kinetics of Thermal Liquefaction of Belle Ayr Subbituminous Coal Donald C. Cronauer,, Yatish T. Shah,' and Raffaele G. Ruberto Gulf Research & Development Company, Pittsburgh, Pennsylvania 15230
This paper presents the results of a kinetic study of thermal liquefaction of Belle Ayr subbituminous coal. The experimental work was carried out in a laboratory-scale, continuous stirred tank reactor. The results for the conversion of coal and the production of pre-asphaltenes, asphaltenes, oils, and gases such as C1-C6, NH3, H2S, CO, COP,and water are given as functions of slurry space time and temperature. A temperature range of 400 to 470 OC and a space time range of approximately 5 to 55 min were examined. All the experimental data were taken at a total unit pressure of 2000 psig and coal-to-solvent ratio of 1: 1.5. Two solvents, hydrogenated anthracene oil and hydrogenated phenanthrene, were investigated. The experimental results were correlated by a kinetic model which assumes the reaction mechanism
/\gases
coal
f
oils
pre-asphaltenes
/
asphaltenes
It is shown that this reaction mechanism (with all reaction rates assumed to be pseudo first order with respect to reacting species) correlates data reasonably well at the temperature levels of 400, 425, and 450 O C for all space times, and at 460 and 470 OC for small space times.
Introduction Recent demands for fuel have increased the importance of coal liquefaction processes. Various processes for coal liquefaction are currently being examined a t the pilot plant level including catalytic [Synthoil, Akhtar et al. (1974a, 1974b); H-Coal, Johnson e t al. (1973); Gulf's CCL] and noncatalytic [Exxon, Furlong et al. (1976) and SRC, Anderson and Wright (1975)l. Very few kinetics data on these processes have been published. A good review of the reported studies on catalytic coal liquefaction has been published by Oblad (1976). The kinetics of liquefaction processes have been examined by several investigators, the most recent studies being those of Guinn e t al. (1975, 1976), Whitehurst et al. (1976), Plett e t al. (1975), Shah et al. (1978), and Reuther (1977). These and similar studies point out a strong need for good kinetic data, which are lacking in most publications. A large amount of published data is obtained either in batch reactors or in pilot-scale demonstration units. These data are, of course, not easily amenable to evaluation of the kinetic rate expressions for the coal liquefaction process. In the present study, we outline a kinetic model for thermal liquefaction of Belle Ayr subbituminous coal carried out in a carefully designed laboratory-scale continuous stirred tank reactor. The effect of space time and temperature on conversion of coal and production of pre-asphaltenes, asphaltenes, and oils is illustrated. A kinetic model which is somewhat more sophisticated than those published in the literature is used to correlate the experimental data. Department of Chemical and Petroleum Engineering,University
of Pittsburgh, Pittsburgh, Pa. 15261.
0019-7882/78/1117-0281$01.00/0
Experimental Section Materials. Subbituminous coal from the Belle Ayr Mine, WY, of Amax Coal Co. was used. The proximate and ultimate analyses of this coal are given in Table I. I t is noted that the ultimate analysis was performed on samples of pulverized coal with subsequent correction for moisture content. Table I. Analysis of Belle Ayr Coal Samples As received Proximate analysis,a Wt % 30.25 0/6 Ash 7.01 % Volatile 28.26 % Fixed carbon 34.48 100.00 Heat of combustion. Btullb 8141 % Moisture
Dry basis
... 10.05
40.52 49.43 100.00 11 671
Ultimate Analysis,b Wt % on a Dry Basis Carbon 69.3 Hydrogen 4.3 Nitrogen 1.0 Oxygen, determined 19.9 Difference 14.5 Sulfur, organic 0.45 Pyritic 0.06 Sulfate 0.00 Total 0.5 Ash, total metals 10.3 5.0 Total' 100.0 Average of two spot samples. Average of samples taken during experimentation. Sum of C, H, N,ODiff, ST, ash =
*
100.0.
0 1978 American Chemical Society
282
Ind. Eng. Chem. Process Des. Dev., Vol. 17, No. 3, 1978
I
Table 11. Analysis of Solvents Designation HA0 HPh Carbon Hydrogen Nitrogen Oxygen suifur
Chemical Analysis, Wt % 91.20 6.86 0.65 1.01 0.28 100.00
"-PENTANE
SAMPLE
1
BENZENE
PYRIDINE
ASPHALTENES PREASPHALTENES PYRIDINE. INSOLUBLES
91.08 8.83 0.01
...
0.08 100.00
P IRID I NE-
ASPHALTENES
Physical Properties Viscosity, cSt at 25 "C Specific gravity (60/60) Insolubles, pentane, wt % benzene, wt %
21.28 1.092 0.16 0.05
Distillation, O 0.p.. wt % 10 50 90 E.p. Residue, wt %
Figure 2. Solvent extraction sequence.
10.80 1.038 0.00 0.00
m e t e r impeller. The agitator was run at sufficient speed (lo00 rpm) to ensure a minimum of mass transfer effects. There were four baffles in the reactor. As a check of the degree of agitation, a glass model was fabricated. Visual observations of mixtures of water, air, and quartz chips were made a t agitation speeds between 500 and lo00 rpm. The degree of mixing was effective and essentially no change was observed over this range. In addition, liquefaction runs were made at 1000 and 1500 rpm; no effect of agitation was observed. Liquefaction runs were started by first bringing the unit to operating pressure and adjusting the hydrogen flow. The reactor was heated to within 50 "C of operating temperature and the feed slurry was started. The runs were of 4-6 h in length. With the exception of the experiments of about 50 min space time, at least six reactor volumes passed through the unit prior to sampling. Samples were taken during the latter two 60-min periods. These products consisted of slurry, aqueous and oil phases from the traps, and gases. The slurry was filtered in a Buchner funnel a t ambient temperature unless steam was necessary. The wet filter cake and filtrate were subjected to the solvent extraction scheme illustrated in Figure 2. From this "parallel" technique, the amounts of the following components were calculated: oil = (total sample) - (pentane insolubles); asphaltenes = (pentane insolubles) - (benzene insolubles); pre-asphaltenes = (benzene insolubles) - (pyridine insolubles); pyridine insolubles = as determined.
C
174 262 335 401 483 1.1
219 270 323 351 390 0.0
The solvents were hydrogenated anthracene oil (HAO) and hydrogenated phenanthrene (HPh). Analyses are given in Table 11. The anthracene oil (Reilly Tar and Chemical Corp.) was hydrogenated over a cobalt molybdenum-type catalyst at a temperature of 413 "C and a pressure of 27.6 M P a (4000 psig). Phenanthrene (Aldrich Chemical Co., Inc.) was catalytically hydrogenated a t similar conditions but a t a reduced pressure of 10.4 MPa (1500 psig). Liquefaction Apparatus and Experimental Procedure. A flow diagram of the coal liquefaction unit is shown in Figure 1. In summary, the unit consisted of a feed tank, Moyno recirculating pump, Hills-McKenna high pressure pump, preheater, a stirred autoclave, let-down valves controlled from unit pressure, receivers, and a gas chromatograph. The reactor was 293 cm3 in volume with an inside diameter of about 7.6 cm. The reactor was equipped with a single large (5.1 cm) diDISPLACEMENT GAUGE
R
4
VALVES
D.T M
PROD.
DRY I C E a G.C
TRAP
K.O. TRAP
WET
ICE TRAP
DRY ICE TRAPS
(2) \\
4
PREHEATER
LIJ S L U R R Y PROD.
RECYCLE
H I G H PRESS, PUMP
7CAKE TO E X T R A C T I O N
E S T E A M FUNNEL D R Y ICE T R A P
FILTRATE
Figure 1. Experimental apparatus.
a VACUUM
TO E X T R A C T I O N
Ind. Eng. Chem. Process Des. Dev., Vol. 17, No. 3, 1978 283
Table 111. Results of Noncatalytic Solvent Runs
Space time, min Temperature, OC Pressure, MPa Carbon Hydrogen Oxygen Sulfur
Solvent
Is
... ... ...
80 375 13.9
Run no. 3s
2s Reactor Conditions 20 375 20.9
20 425 20.9
Liquid Analysis, Wt % 91.49 91.61 6.95 6.93 1.11 1.08 0.26 0.22
91.20 6.86 1.01 0.28
4s
5s
80 425 20.9
80 450 20.9
...
91.55 6.78 1.01 0.23
6s 10 425 20.9
91.38 6.97 1.00 0.18
6.80 1.02 0.22
91.76 6.91 1.18
0.23
Table IV. Summary of Belle Ayr Liquefaction Runs with HPh Solvent
Space time, min Temperature, "C Percent solvation Oil yield (g/lOO g of MAF) Feed, g/h Coal, MAF Coal, ash Coal, moisture Coal total Solvent total Hydrogen consumption Total Products
co coz c1-c3 c4-c6
HzO (out) NH3 plus HzS Oils Asphaltenes Pre-asphaltenes Pyridine insolubles Total
Run no. 5
1
2
3
4
6
7
8
9
5.1 400 51.37 4.24
26.4 400 72.33 30.87
23.8 400 71.44 33.01
46.4 400 79.15 44.61
7.0 450 80.47 33.84
12.6 450 89.90 41.26
18.9 450 91.48 41.88
30.2 450 98.6 70.79
53.8 450 98.05 66.82
979.25 104.68 400.90 1484.83 2227.25 2.06 3710.02
191.21 20.44 78.28 289.94 434.90 1.33 726.17
211.56 22.62 86.61 320.79 481.19 0.97 802.94
108.66 11.62 44.49 164.76 247.14 0.62 412.52
728.14 73.41 295.71 1097.26 1645.89 0.22 2742.93
399.04 42.66 163.37 605.06 907.59 0.76 1513.41
269.00 27.12 109.25 405.37 608.05 1.71 1015.12
168.35 16.97 68.37 253.70 380.55 3.93 638.18
93.82 10.03 38.41 142.27 213.40 1.97 353.70
4.81 36.98 6.86 11.25 412.86 0.40 2268.72 129.47 257.66 580.93 3709.94
1.02 13.74 3.17 3.03 90.64 0.63 493.92 32.33 14.35 73.34 726.17
0.07 10.08 2.38 2.36 101.20 1.04 551.03 28.02 22.92 83.04 802.95
0.51 8.53 1.84 2.22 48.86 0.96 295.61 11.01 8.45 34.27 412.57
3.30 36.42 12.98 13.05 314.27 2.19 1892.26 129.50 123.36 215.58 2742.90
3.42 26.59 16.76 11.48 188.07 3.39 1072.23 64.11 44.38 82.98 1513.40
2.42 19.07 14.09 9.81 128.25 1.63 720.71 41.73 27.38 50.05 1015.12
1.02 5.96 6.88 5.00 90.43 2.02 499.72 8.31 3.21 15.63 638.18
1.58 15.52 1.58 0.77 42.26 0.31 276.08 2.64 1.10 11.86 353.70
Table V. Summary of Belle Ayr Liquefaction Runs with H A 0 Solvent
Space time, min Temperature, "C Percent solvation Oil yield (g/lOO g of MAF) Feed, g/h Coal, MAF Coal, ash Coal, moisture Coal total Solvent total Hydrogen consumption Total Products
co COZ c1-c3 c4-c6
Hz0 (out) NH3 plus HzS Oils Asphaltenes Pre-asphaltenes Pyridine insolubles Total
5.6 400 55.69 8.92
15.1 400 66.29 17.98
20.3 400 58.94 17.79
48.4 400 74.74 33.94
5.8 425 63.71 11.95
425 67.14 14.73
27.9 425 77.65 34.49
903.98 91.13 367.12 1362.24 2043.36 5.08 3400.52
334.80 35.79 137.07 507.65 761.48 0.60 1269.73
250.57 25.26 101.76 377.60 566.40 0.90 944.89
104.88 10.57 42.59 158.05 237.08 0.63 395.76
867.76 87.48 352.41 130.66 1961.49 2.04 3266.54
452.97 45.67 183.96 682.59 1023.89 1.57 1704.91
181.96 18.34 73.90 274.20 411.30 2.38 687.88
2.72 56.61 4.77 9.33 356.58 1.73 2123.97 225.49 127.69 491.73 3400.62
0.28 26.72 4.96 10.92 158.71 1.32 821.66 55.00 41.48 148.65 1269.68
1.40 20.80 3.57 7.03 106.21
0.13 6.99 0.37 3.58 48.50 .54 272.69 20.71 4.90 37.35 395.76
3.38 45.93 8.83 14.93 363.20 2.27 2065.17 224.55 135.65 402.42 3266.31
3.22 43.09 11.16 14.96 195.04 1.92 1090.60 78.01 72.40 194.51 1704.91
1.54 15.79 7.17 7.99 80.30 1.05 474.06 26.24 14.74 59.00 687.89
.oo
610.81 48.86 18.42 128.16 945.25
11.2
284
Ind. Eng. Chem. Process Des. Dev., Vol. 17, No.3, 1978
Table VI. Summary of Belle Ayr Liquefaction Runs with H A 0 Solvent
Space time, min Temperature, "C Percent solvation Oil yield (g/100 g of MAF) Feed Coal, MAF Coal, ash Coal, moisture Coal total Solvent total Hydrogen consumption Total Products
co
CO2
c1-c3 c4-c6
HzO (out) "3 plus H2S Oils Asphaltenes Pre-asphaltenes Pyridine insolubles Total
17
18
19
20
21
22
5.1 450 67.22 14.48
11.7 450 81.48 29.69
16.7 450 83.88 26.33
24.8 450 89.38 37.98
24.9 450 79.77 25.66
39.5 450 87.93 22.36
993.09 100.12 403.31 1496.51 2244.77 0.32 3741.61
432.95 43.65 175.83 652.43 978.64 0.44 1631.51
302.33 32.32 123.77 458.42 687.62 8.82 1154.86
207.21 20.89 84.15 312.25 468.38 0.02 780.62
203.78 20.54 82.76 307.08 460.62 1.72 769.42
127.69 13.65 52.27 193.61 290.42 1.88 485.90
1.56 21.60 6.71 8.33 442.95 6.05 2388.56 295.44 145.65 425.68 3742.52
3.11 35.19 17.18 10.03 193.87 2.30 1107.21 94.20 44.58 123.84 1631.50
2.40 26.14 18.47 16.59 138.50 2.39 767.24 61.73 40.36 81.04 1154.87
1.01 8.60 7.13 4.03 90.38 1.05 547.08 66.42 12.02 42.89 780.61
1.45 11.63 8.87 7.22 94.53 1.16 512.91 44.66 25.22 61.77 769.40
0.99 9.59 11.38 6.33 64.58 1.42 318.96 28.06 15.52 29.06 485.91
23
24
25
26
27
29.0 450 79.50 31.52
5.5 460 72.73 17.86
11.5 450 77.88 25.20
5.2 470 75.50 8.66
10.6 470 83.15 22.39
103.50 10.43 42.03 155.96 233.95 1.43 391.34
923.78 93.13 375.16 1392.08 2088.12 1.35 3481.54
440.03 44.36 178.70 663.09 994.63 1.61 1659.33
971.58 97.95 394.58 1464.11 2196.17 4.78 3665.06
477.84 48.17 194.06 720.07 1080.11 5.88 1806.06
0.73 4.42 6.50 4.11 50.53 0.73 266.57 18.52 7.56 31.65 391.32
8.91 38.89 17.37 16.00 424.53 3.77 2253.09 165.47 208.46 345.04 3481.54
1.25 11.19 7.50 7.35 213.88 2.01 1105.52 145.33 23.61 141.68 1659.33
9.06 75.62 42.70 43.31 411.98 3.39 2280.27 282.34 180.35 366.02 3665.03
5.08 36.03 30.84 22.95 200.78 3.70 1187.09 103.40 87.48 128.70 1806.07
~~
Space time, min Temperature, "C Percent solvation Oil yield (g/100 g of MAF) Feed Coal, MAF Coal, ash Coal, moisture Coal total Solvent total Hydrogen consumption Total Products
eo
coz
c1-c3 C4-cS HzO (out) NH3 plus H2S Oils Asphaltenes Pre-asphaltenes Pyridine insolubles Total
Solvent Stability. Some experiments were performed to test the stability of the solvent, HAO, a t reaction conditions. As shown in Table 111, the chemical analyses of the reaction products remained essentially unchanged a t reaction temperatures up to 450 "C. The distillation curves of the products also overlapped, showing only minor changes. Results The liquefaction experiments were carried out a t five temperature levels, namely 400,425,450,460, and 470 "C, and a t space times ranging from approximately 5 to 55 min. The total reactor pressure was kept constant a t 13.8 MPa (2000 psig). Since, from a practical standpoint, high coal/solvent ratios are desirable, the experiments were performed a t a coal/solvent ratio of 1:1.5. Two solvents, namely hydrogenated anthracene oil and hydrogenated phenanthrene (designated H A 0 and HPh, respectively), were studied. Some of the ex-
~~
periments were duplicated. The coal solvation results obtained in these duplicate runs differed by less than 3%. A summary of the material balances is shown in Tables IV, V, and VI. Kinetic Model The data obtained in this study were correlated by a simple kinetic model. The kinetic model assumes the reaction mechanism
' . gases
Kg/
KO
oils t--coal
\ +
Kg
J
asphaltenes
pre-asphaltenes
(1)
Ind. Eng. Chem. Process Des. Dev., Vol. 17, No. 3, 1978
The above reaction mechanism assumes that coal initially disintegrates and produces pre-asphaltenes, asphaltenes, oils with ranging boiling points, and gases (such as Cl-Cs, water, ammonia, hydrogen sulfide, etc.). Asphaltenes and pre-asphaltenes undergo a further series reaction, namely, preasphaltenes asphaltenes oils. The above mechanism is similar to those reported in the literature, except that gases and some of the asphaltenes and oils are assumed to be formed directly from coal. Typical literature kinetic mechanisms assume strictly consecutive types of reactions, i.e., coal pre-asphaltenes oils. One or more reactions consume hydrogen. We assume dissolved hydrogen to be in excess. The present data indicate that approximately 60% of the total hydrogen consumed is used to produce gas. The kinetic model for the reactor is developed using a carbon balance which excludes ash and moisture in the coal. It is also assumed that coal contains some unconvertibles (fusinite, etc.). The amount of unconvertibles (Le., equilibrium coal concentration) depends upon the temperature and the nature of the solvent. The net increase in the weight of slurry phase due to hydrogen consumption is small and, therefore, neglected.
-
-
+
+
285
The coal balance around the isothermal stirred tank reactor is described by the dimensionless equation
%)
n K r (1 \ Lei/ 1 - L0 = (2) Cci l+Kr where K = K, K, K, K Oand C,i and Cc, are the inlet and outlet moisture- and ash-free weight fractions of coal in the slurry. C, is the equilibrium weight fraction of coal (i.e., weight fraction of coal remained in the slurry at large space time). Its value is assumed to depend upon temperature and space time. K,, K,, K,, and K Oare the rate constants for the various reactions shown in eq 1. r is the space time in the reactor. It is defined as pV(1 - cg)/rn.Here Vis the volume of reactor, cg is the gas holdup in the reactor, rn is the mass flow rate of the coal-oil slurry, and p is the density of coal-oil slurry a t ambient conditions ( 2 5 "C). The gas holdups under various reaction conditions were estimated from the available literature (Bates et al., 1966; Calderbank, 1967; Sideman et al., 1966). Similar mass balances for other reacting species can be expressed as in eq 3-6.
+ +
+
Gases: HA0
400'C 0 HA0 450'C A HPh 450'C Model Predict l o i s Reactor Conditions Pressure 13 8 MPa i 2 O O O p s ~ g I Coai/Solvent = I / 1 5 0
---
Pre-asphaltenes:
Asphaltenes:
Cci
I 20
0
S l u r r y Space Time l m l n
40
60
,
HA0
400OC
400'C 450T
0
_ _ _ M o d e l Predlctlons Reactor Conditions Pressure 13 8 MPa ~ 2 0 0 0 p s ~ g I , Coa~/Splvent= 1 / 1 5 ,
li
Oils:
I
Figure 3. Coal solvation vs. space time: model predictions and experimental data.
0 HPh A HPh
1+Kr
,
, I
'i
Cci
l+Kr
Here C,,, C,,, and Cooare the reactor outlet weight fractions of the pre-asphaltenes, asphaltenes, and oils, respectively, and C,i is that of the inlet oil concentration. K,, and Ka, are the askinetic constants for the reactions pre-asphaltenes phaltenes and asphaltenes oils, respectively. +
+
i
at 400'C
400'C
60 Slurry SpaceTime Imin 1
Figure 4. Net production of pre-asphaltenes vs. space time: model predictions and experimental data.
Discussion From the experimental data obtained at the above temperatures, the values of the rate constants K,, K,, Kpa,K,, K,,, and K Owere obtained by fitting eq 2 through 6 to the experimental data by nonlinear regression analysis. The kinetic constants at 460 and 470 "C may be somewhat inaccurate because these were obtained from the experimental data a t small space times only. As shown in Figures 3 to 7, for some typical reaction conditions, the predicted curves for the conversion of coal, and the productions of pre-asphaltenes, asphaltenes, oils and gases vs. space time agree reasonably well with the experimental data for both coal-solvent systems.
286
Ind. Eng. Chem. Process Des. Dev., Vol. 17, No. 3, 1978 HA0 400'C HPh 400'C HPh 450% _ _ _ M o d e l Predictions Reactor Conditions: Pressure: 13 8 MPa ( 2 0 0 O p s 1 g I Coal/Solvent- 1 / 1 5 0
I-A, ----I
I
,
02J
I
0 0
HA0 HA0
A
HPh
4OOOC
450OC 450% Model Predictions Reactor Conditions: Pressure: 13 8 MPa 12000prigl Cosl/Solvent= 1/1.5
0
___
A
~
I
0:
'.".
," /
I
,
,#&------
0
I/ ,"I
0
02
31: \
I
\
I
,
\
I
A \
01
\'
L
20
0
40
Slurry Space Time ( m i n I
Figure 5. Net production of asphaltenes vs. space time: model predictions and experimental data.
- - - Model Predlctlons Condltlons Pressure 13 8 MPa Reactor
20
40
S l u r r y Space Time 1min.I
Figure 7. Gas production vs. space time: model predictions and experimental data. (2000prig I
Coal/Solvent = I / 1 5
I
1 AI'
07t
I
/
0 HA0 0 HPh
A HPh
I
400'C 400'C 450%
I
/
/ /
/
0.5
/
I
t S l u r r y Space Time Irnin.1
Figure 6. Net production of oils vs. space time: model predictions and experimental data.
10''
135
14
15
145
lnveraa Absolute Temperature 1 /T x lo3 ('K.'I b
Figure 8. Arrhenius plots for the rate constants: HA0 solvent. Arrhenius plots for the rate constants for the coal-HA0 and coal-HPh systems are shown in Figures 8 and 9, respectively. The activation energies and frequency factors for various reactions are listed in Table VII. These activation energies appear to be somewhat low for a purely thermal reaction. Based on the agitation study and liquefaction experiments in other units of our laboratory, these low activation energies are not a result of diffusion resulting from questionable agitation. I t is postulated that they may be low as a result of donor solvent having to diffuse into coal and asphaltene-type gel particles. This diffusion is an essential part of the liquefaction process and not directly affected by agitation. In further confirmation, the activation energies of the coal-related reactions were lower than those of asphaltene and pre-asphaltene reactions (see Table VII). I t is also considered that the activation energies may be low as a result of grouping many individual reactions (including possible reverse reactions) into the simple mechanism given in eq 1. Earlier studies [Oblad (1976), Guin et al. (1975, 1976), Neavel (1976)l have shown that in the temperature range of
350-450 O C , degradation of coal (particularly bituminous coal) occurs to an extent sufficient to solubilize the coal within about 1min. The fragments break from larger molecules by thermal cleavage and the resulting segments are free radicals. These fragments may be stabilized by reacting with each other or with a hydrogen donor solvent. If these stabilization reactions include hydrogen transfer, lighter oil products are formed. Otherwise, in general, heavy asphaltene or pre-asphaltene types of materials are formed. During these fragmentation reactions, the light gases are either stripped from heteroatom groups or they are formed by cracking of alkyl groups. The magnitudes of the rate constants K,, K,, K,, and KO indicate the importance of the direct coal conversion reactions to gases, pre-asphaltenes, asphaltenes, and oils, respectively. From Figure 8, it is interesting to note that for the coal-HA0 system, the kinetic constants for the direct reactions are about the same order of magnitude. Furthermore, a t high temper-
Ind. Eng. Chem. Process Des. Dev., Vol. 17, No. 3, 1970
287
Reactor Conditions Pressure 13 8 MPa l 2 0 0 O p s i g I Coal/Soivent = 1/ 1 5
P
01-
+
g
U
4
L
650
675
700 Temperature
725
750
(OK1
Figure 10. Calculated equilibrium liquefaction concentration of moisture-ash-freecoal as a function of temperature. 10
135
1 4
145
Inverse Absolute Tarnperslurs 1 /T x 10’
1 5
(OK”)
Figure 9. Arrhenius plots for the rate constants: HPh solvent.
Table VII. Activation Energies and Frequency Factors Activation energy, cal/g-mol CoalCoalHA0 HPh
Reaction
---
Coal oils Coal preasphaltenes Coal asphaltenes Coal gases Coal oils pre-asphaltenes asphaltenes gases Pre-asphaltenes asphaltenes Asphaltenes oils
-
-
Frequency factor, min-l CoalHA0 Coal-HPh
lo3 2.1 X los lo3 4.94
14 100 28 900 13 800 4 300
3.11 X 2.81 X
15 600 8 600 21 500 10 500 16 700 20 500
1.12 X lo4 9.63 X lo1 8.72 X lo5 3.85 X lo2 9.15 X lo4 1.59 X lo6
12 800 33 900
9.66 X lo2 2.48 X lo9
16 000 25 600
1.42 X lo3 1.53 X lo7
-
-
atures, the reactions of coal gases and coal asphaltenes are more favorable. The magnitude of K,, and K,, indicate the importance of hydrocracking reactions pre-asphaltenes asphaltenes and asphaltenes oils. The data shown in Figure 8 indicate that for the coal-HA0 system, the first hydrocracking reaction (pre-asphaltenes asphaltenes) occurs favorably, but the second reaction occurs at a very slow rate. This means that in the coal-HA0 system, the major oil product is formed directly from coal. In principle, the hydrogen donor solvent required to produce oil products during coal liquefaction can come from coal (Neavel, 1976). In the present study, hydrogenated phenanthrene was used to see the effect of an externally supplied effective hydrogen donor solvent on the distribution of the products. As shown in Figure 10, the equilibrium concentrations of unconverted coal in the case of hydrogenated phenanthrene are smaller than those with hydrogenated anthracene oil. This means that some of the so-called unconvertibles in the case of the coal-HA0 system were converted in the presence of HPh. The nature of the solvent thus plays a very important role on the extent of coal liquefaction.
-
-
-
Various rate constants of the coal-HPh system described in Figure 9 indicate that the values of KOand K,,, particularly a t high temperatures, can be significantly increased with the use of a good hydrogen donor solvent. Therefore, the production of oil is increased with the use of a good hydrogen donor solvent. A large number of coal liquefaction studies have been made using hydrogen donor solvents such as tetralin. With the hydrogenated anthracene oil used in this study, the identity and nature of the hydrogen donor species are less clear and their concentrations are moderate to small. At high temperatures and space time when a very large number of coal fragments (or free radicals) are being produced, the concentration of hydrogen donor compounds becomes important. As observed, hydrogenated phenanthrene at these conditions gives more favorable yields than hydrogenated anthracene oil. For the production of stable and desirable product yields, the need for sufficient amounts of hydrogen donor solvent is well recognized. In a catalytic process, the catalyst is supposed to hydrogenate the solvent. A more sophisticated kinetic model which takes into account the effects of solvent to coal ratio, pressure, and catalyst should involve the role of hydrogen donor solvent in one or more reactions in eq 1. Conclusions A reaction model using series and parallel reaction paths has been developed to describe the liquefaction of Belle Ayr subbituminous coal in hydrogenated anthracene oil and hydrogenated phenanthrene. It has been shown that in the temperature range of 400 to 470 “C, this model correlates reasonably well the distributions of pre-asphaltenes, asphaltenes, and oils in the product. The present experimental results indicate that the production of oil from coal will be substantially increased with the use of good hydrogen donor solvent. Acknowledgment For partial financial support, the authors are grateful to the Electric Power Research Institute. We wish also to acknowledge our indebtedness to W. C. Rovesti, R. G. Goldthwait, D. M. Jewell, and H. G. McIlvried for helpful discussions, and to K. A. Kueser for the handling of experimental data. Literature Cited Akhtar, S., Friedman, S., Yavorsky, P. M.. AIChE Symp. Ser., 70 (137), 106 (1974a). Akhtar, S.. Lacey, J. J., Weintraub, M., Resnik, A. A., Yavorsky, P. M., Paper No. 356, presented at 67th Annual AlChE Meeting Dec 1-5, 1974, Washington, D.C., 1974b.
Ind. Eng. Chem. Process Des. Dev., Vol. 17, No. 3, 1978
288
Anderson, R. P.,Wright, C. H., Am. Chem. SOC.,Div. Fuel Chem., Prepr.. 20 (I),
2 (1975). Bates, R. L., Fondy, P. L., Fenic, J. G., in "Mixing I", p 11 1. V. W. Uhl and J. E. Gray, Ed., Academic Press, New York, N.Y., 1966. Calderbank, P. H.. in "Mixing 2",p 1, V. W. Uhl and J. E. Gray, Ed., Academic Press, New York, N.Y., 1967. Furlong, L. E., Effron, E., Vernon, L. W., Wilson, E. L., Chem. Eng. Prog., 72 (8).
69 (1 976). Guin, J. A., Tarrer, A. R., Taylor, Z. L., Green, S.C., Am. Chem. Soc.. Div. f u e l Chem., Prepr., 20 (l),66(1975). Guin, J. A.. Tarrer, A. R.. Pitts, W. S.,Prather, J. W., Am. Chem. SOC.,Div. Fuel Chem., Prepr., 21 (5),170 (1976). Johnson, C. A., Wolk. R. H.,Chervenak, M. C., Johanson, E. S.,C b m . Eng. prog., 69 (31.52 119731. - -, Neavel, R. C., Fuel, 55, 237 (1976). Oblad, A. G., Catal. Rev. Sci. Eng., 14 ( I ) , 83-96 (1976). \
~
Plett, E. G., Alkidas, A. C., Roger, F. E., Mackiewicz, A. Z., Summerfield, M., paper presented at University-ERDA Contractor Conference, Salt Lake City, Utah, Oct 22-23, 1975. Reuther, J. A,, lnd. Eng. Chem. Process Des. Dev., 18, 249 (1977). Ruberto, R. G., Jewell, D. M.. paper presented at the NSF Workshop, "Analytical Needs of the Future as Applied to Coal Liquefaction", Greenup, Ky., 1974. Shah, Y. T., Cronauer, D. C., Mcllvried, H. G., Paraskos, J. A,, lnd. Eng. Chem. Process Des. Dev., accompanying article in this issue, 1978. Sideman, S., Hortascu, O., Fulton, J. H., lnd. Eng. Chem., 58 (7),32 (1966). Whitehurst, D. D., Farcasiu. M.. Mitchell, T. O., EPRl Annual Report No. AF-252, RP 410-1, (Feb 1976).
\
Receiued for review April 29, 1977 Accepted February 2,1978
Kinetics of Catalytic Liquefaction of Big Horn Coal in a Segmented Bed Reactor Yatish T. Shah,*
' Donald C. Cronauer, Howard G. Mcllvried, and John A. Paraskos
Gulf Research & Development Company, Pittsburgh, Pennsylvania 15230
The kinetics of liquefaction of Big Horn subbituminous coal is experimentally studied in a Gulf patented segmented bed catalytic coal liquefaction (CCL) reactor. The reaction mechanism is assumed to be
by-products
1
c,-c,
- 1I / coal
light gases
\
water
400-650 "F (204-343 OC)
furnace
-
C,-400 (C,-204 , _
O F
"C)
naphtha
Oil
650
O F +
( 3 4 3 "C+)
heavy fuel oil
In the present model all reactions from coal are assumed to be irreversible and first order. The hydrocracking steps are assumed to be either first or zero order. Based on the aforementioned reaction mechanism, the effects of (a) slurry space time, (b) reactor temperature, (c) reactor pressure, (d) initial coal concentration, and (e) gas flow rate on the product distributions are examined both experimentally as well as theoretically. Slurry space time and the reaction temperature are found to have a pronounced effect on the CCL product distribution. Over the ranges considered, reactor pressure, feed coal concentration, and the gas flow appeared to have relatively mild effect on the product distribution.
Introduction Because of a potential major impact on our future energy supply, the production of oil by coal liquefaction is being extensively evaluated. Various processes, such as SRC, Exxon, Consol, Synthoil, H-coal, and the Gulf process, are being evaluated a t the bench- or pilot-scale. In most coal liquefaction processes, coal is liquefied in the presence of a solvent. In addition, the coal-oil slurry may be contacted by a hydrogen-rich gas, and a catalyst may also be present. The process of liquefaction produces a wide boiling range liquid, as well as some light gases (Cl-C,) and byproducts, such as water, ammonia, hydrogen sulfide, etc. Some data for the kinetics of coal liquefaction have been published (Curran et al., 1966; Plett et al., 1966; Guin et al., 1966; Fu and Batchelder, 1976; Cronauer et al., 1978). The reported data were mostly obtained in bench-scale reactors. Both thermal and catalytic liquefaction processes have been To whom all correspondence should be sent at the Department of Chemical Engineering,University of Pittsburgh, Pittsburgh, Pa. 15261.
examined. Guin et al. (1976) studied the mechanism of coal particle dissolution and Neavel (1976),Kang et al. (1976), and Gleim (1976) examined the role of solvent on coal liquefaction. I t is believed that hydrogen donor solvent plays an important role in the coal liquefaction process. This role was recently investigated by Reuther (1977) and the reaction paths in a donor solvent coal liquefaction process have been reviewed by Squires (1976). The effects of coal minerals on reaction rates during coal liquefaction have been examined by Tarrer et al. (1976), whereas the short contact time coal liquefaction process has been investigated by Whitehurst and Mitchell (1976). A review of the reported studies has recently been given by Oblad (1976). The primary purpose of this paper is to describe the kinetics of coal liquefaction in a pilot-scale unit. The specific coal studied was Big Horn subbituminous coal. The experimental data were obtained in a pilot-scale Gulf patented (Chun et al., 1976) reactor. Some of the data reported here were recently evaluated by a simple kinetic model (Shah et al., 1978).In this paper we evaluate a larger variety of data by a more detailed and sophisticated kinetic model.
0019-7882/78/1117-0288$01.00/0 0 1978 American Chemical Society
