Literature Cited
Gast, Th., Vahuum-Technih 14, 41-4 (1965). Jost, W., “Diffusion in Solids, Liquids, Gases,” pp. 64, 292, Academic Press, New York, 1962. Meier, E., “Trocknungsverlauf hydroskopischer Stoffe mit konzentrations- und temperaturabhangigen Diffusionskoeffizienten am Beispiel der Perlontrocknung,” dissertation, Aachen Technical University, 1968. Robens, E., Lab. Praht. 18, 292-314 (1969). Robens, E., Robens, G., Sandstede, G., Vacuum 13, 1037 (1963).
Robens, E., Sandstede, G., J . Sci. Instr. 2, Ser. 2, 3658 (1969). RECEIVED for review March 10, 1969 ACCEPTED July 31, 1969 Abridged version of a lecture given by Erich Robens before the VDI-Fachgruppe Kunststofftechnik, Dusseldorf, Germany, Kov. 21, 1969. Extensive paper submitted for publication in Kunststofe (in German). Investigations conducted a t BattelleInstitut e.V., Frankfurt (Main), on behalf of a r t h u r Pfeiffer Hochvakuumtechnik GmbH, Wetzlar.
THERMAL CRACKING OF LOW-TEMPERATURE LIGNITE PITCH R I C H A R D
1.
RICE,
DELMAR
R.
F O R T N E Y ,
A N D
JOHN
S .
B E R B E R
Morgantou’n Coal Research Center, Bureau of Mines, U .S . Department of the Interior, Morgantoun, W . Va. 26X5
Low-temperature lignite pitch was thermally cracked in a 4-inch i.d. reactor at 1200” to 1450°F. and pitch feed rates of 5.5, 7.0, and 9.5 pounds per hour to produce coke, oil, and gas. Temperature had a much greater influence on product yields than pitch feed rate. Highest coke yield ( 3 0 weight “10)was obtained a t minimum pitch feed rate and 1450“ F. Maximum oil yield ( 2 8 weight Yo) was obtained at the lowest feed rate and intermediate temperature ( 1 3 2 5 ° F . ) . Gas yield was highest ( 3 0 weight YO) a t the lowest feed rate and maximum temperature ( 1450” F.).
THERMAL cracking of pitch
from the carbonization of Texas lignite a t low temperature has been under investigation as a means of producing aggregate and a binder that could be used to fabricate carbon metallurgical electrodes. A previous report covered preliminary tests on the thermal cracking of lignite pitch a t 1450”F. in a 2 1 2 -inch-diameter reactor and showed that a variety of products could be obtained (Berber et al.: 1965). In this work, a 4-inch-diameter reactor was utilized t o evaluate oil, coke. and gas quality and yields as a function of pitch feed rate and temperatures between 1200” and 1450”F. Equipment and Materials
Figure 1 is a flow sheet of the system. The cracking unit consisted of a 93-inch length of 4-inch, schedule 40, Type 304 stainless steel pipe, heated electrically. Total heating capacity of the cracker was 13.54 kw. A 1-inch pipe extended up through the center to within about 212 feet of the top and was perforated with I I-inch openings to allow withdrawal of gas and oil vapors from the reaction zone. Pitch utilized as feed material was prepared by distilling crude low-temperature lignite tar under vacuum to an atmospheric boiling point of 660°F. The pitch yield was about 45‘; of the tar. Ultimate analysis and physical properties of the pitch are summarized in Table I.
30
Ind. Eng. Chern. Prod. Res. Develop., Vol. 9, No. 1, March 1970
Vent
r
Sample
1
Thermal cracker
tf
/Scrubber
Cooling water
u
Figure 1. Thermal cracking system
Table I. Physical and Chemical Properties of Feed Pitch
IJtirnate Analysis,
As Received or Dry'
(5
Carbon Hydrogen Nii,rogen Oxygen Sultur Chlorine
"
Flash point. ' F . Ash, ( c Ductility, cm. a t 77" F. Free carbon Distillation 'To 5 7 2 ~F., Softening point of residue (R and B ) , ' F. Sulfonation index of distillate to 572" F. Water. = 0.00.
84.72 8.53 0.87 4.62 0.90 0.01 510 0.35 0 20.85
6.40 194
0
Procedure
Pitch heated to 400" F. was pumped from the feed tank by a gear pump through electrically heated lines into the top of the thermal cracker and .was analyzed each time for C and H content. No significant difference was found. A flow of' purge gas (115 CO?, 88% Nz)swept products from the reaction zone. Cracked pitch was collected in the receiver and the oil was condensed and separated. Gas was passed through the scrubber and gas meter, then sampled and vented. After each run, which lasted l!? hours, the pitch flow was stopped and the pump was flushed with tar distillate fraction (from tank) to keep the pump from freezing during shutdown. I t is not likely that true steady state was achieved, because of change in reactor geometry by buildup of coke. Heat transfer changed for the same reason but steady-state conditions are believed t o have been approached. After the cracking unit cooled, it was opened a t the top and bottom, and the coke was removed from the walls. Cracked pitch was removed from the receiver and oil was drained from t.he condenser and knockout. Each of the three products was weighed to the nearest 0.1 pound. The oil was distilled under vacuum to an atmospheric boiling point of 752"F., giving about 20 to 305; distillate and 70 to 80% residue. This residue was tested for carbon and hydrogen content and softening point. If the test resu1.t~were in the desired range, the residue was used as an electrode binder. The distillate was oxidized to phtha1.i~and maleic anhydrides or separated into acids, bases, and neutral oils. The neutral oils were separated into n-olefins, paraffins, and aromatics. Gas produced by the thermal cracking was analyzed by gas chromatography.
Softening point (R and B) glycerol, O F. Softening point (cube in glycerol), F. Penetration a t i'i3F., 100 grams, 5 sec. Specific gravity, 79" F./7gCF Bitumen, soluble in C S I Conradson carbon, %
0
1.128 78.80 20.81
30
24
Pitch r a t e I b / h r
0 u
a-70
22
Thermal cracking was carried out a t temperatures from 1200" t o 1450°F. and pitch feed rates varying from 5.5 to 9.5 pounds per hour. A residence time (liquid basis) of nearly 0.7 second was found necessary to crack the pitch. I n initial tests with a 5-foot-long cracker, the residence time was only slightly more than 0.55 second and no cracking occurred.
221
Addition of a 292-foot length of pipe to the cracker increased the residence time to 0.68 second, an increase sufficient to crack the pitch. Figure 2 shows the coke yield for three feed rates a t the temperatures investigated. Highest coke yields were obtained a t the lowest feed rate, indicating that the lower space velocity (gas basis) led to a greater percentage of the pitch coming into contact with the hot wall of the cracker. The higher cracking temperatures also produced more coke. I n the 21/L-inch-diameter reactor that had been used before, coke yields were lower and cracked pitch yields were higher. The oil yield (Figure 3) was higher a t the 5.5 pounds per hour pitch rate, but a yield inversion occurred between 7.0 and 9.5 pounds per hour. Thus, the oil yield decreased with increase in feed rate to a certain point, then increased. The oil may be derived by two means during the cracking process: distillation of feed pitch, and cracking, with the latter predominant a t low rates and giving way to distillation a t high rates. This seems to be verified by the decrease in carbon-hydrogen ratio of the oil residue with increase in feed rate, as shown in Figure 4. Figure 4 also shows that the carbon-hydrogen ratio of the oil residue increases with increasing temperature, indicating the greater cracking effect of higher temperatures.
w
Results and Discussion
194
1,200
A-55
1,300
1,400
1,500
TEMPERATURE,'f
Figure 2. Coke yield as a function of pitch rate and temperature Ind. Eng. Chem. Prod. Res. Develop., Vol. 9,No. 1, March 1970
31
J
+
E
30
E
28
u
i
%
9
1
26
Y P i t c h r a t e , Ib/hr
Y
z 12
9.5 a - 7.0
m 0 m
5.5
5
0-
A-
1
20 1,200
I
I
1,300
1,400
I
Pitch rate rate, Ib/hr
0-95
[L [L
*-70
22
A-5.5 A -55
11 -
I 10 1200
1300
TEMPERATURE, "F
1,400
20 1,200
1,500
Figure 4. Carbon-hydrogen ratio of oil residue
Figure 5 . Effect of pitch rate and temperature on gas yield
Table II. Material Balance for Thermal Cracking of Low-Temperature lignite Pitch
Crude Pitch Rate, PPH
Coke Rate, PPH
9.0 7.0 9.2 6.7 9.5 5.6
2.5 1.9 2.5 1.8 2.7 1.6
1
1
I
1
I
I
800
900
1,000
1 100
1,200
1,300
Rate PPH
Cracked Pitch Rate, PPH
Rate, PPH
Total. PPH
Pph
2.5 1.6 2.6 1.8 2.7 1.6
0.8 0.9 0.9 0.7 0.9 0.2
2.6 2.0 2.6 2.2 2.7 1.7
8.4 6.4 8.6 6.5 9.0 5.1
0.6 0.6 0.6 0.2 0.5 0.5
Oil
1 1,400
TEMPERATURE, "F
Figure 6. Effect of temperature on ethylene-ethane ratio Ind. Eng. Chern. Prod. Res. Develop., Vol. 9,No. 1, M a r c h 1970
1,500
data to give a curve covering a range from 750" to 1450"F. (Figure 6). A similar attempt was made with the methanehydrogen ratio (Figure 7 ) but, as seen, a discontinuity exists a t 1200"F., the separation temperature between the delayed coking and thermal cracking work.
Gas yields (Figure 5) appear to have been less affected by feed rate up to 1300"F., when the yields leveled out a t different percentages; the leveling plateau was higher a t lower feed rates. This indicates that higher cracking temperatures and longer residence time tend to produce more gas because of the greater amount of cracking that occurs. Typical material balances are given in Table 11. Product recovery was generally greater than 90%;. Some losses were incurred in removing the coke from the reactor and the cracked pitch from the receiver. Material balances ranging from 91.6 to 99.1% had been established for the ' system in previous experiments (Berber et al., 1967). Attempts to correlate the ethylene-ethane and methanehydrogen ratios with feed rate were unsuccessful, but correlations of these ratios with temperature were as expected. Ethylene-ethane ratios for 750" to 1200" F., derived from previous work in delayed coking (Berber et al., 19681, fitted in perfectly with the more recent thermal cracking
750
1,400
TEMPERATURE, "F
TEMPERATURE "F
Figure 3. Effect of pitch rate a n d temperature on oil yield
32
1300
~
1,500
Gas
Loss ' I
6.7 8.6 6.5 3.0 5.2 9.0
z W u
h
Data from Berber et al. 119681
0 CT
n 5 2 L
1
k0 yi
z 4
I kW
I
Recent data
t t-
01
750
I
1
1
I
I
800
900
1,000
1,100
1,200
I
I
1,300
1,400
1,500
TEMPERATURE, "F
Figure 7. Effect of temperature on methane-hydrogen ratio I t appears that these ratios are controlled almost entirely __ by temperature, and that pitch feed rate has little effect. This would be expected, because the equilibrium coefficient would react to the temperature of the gases. Su m ma ry
Thermal cracking of low-temperature lignite pitch produced materials that may have a commercial value. Product yields were not greatly affected by pitch feed rate and temperature over the range investigated, but temperature had a noticeable effect on gas composition.
Literature Cited
Berber, J. S., Rice, R. L., Fortney, D. R., IND. ENG. CHEM.PROD. RES. DEVELOP. 6, 197 (1967). Berber, J. S., Rice, R . L., Lynch, R . L., IND.ENG.CHEM. PROD.RES. DEVELOP. 7, 270 (1968). RECEIVED for review May 6, 1969 ACCEPTED August 21, 1969 Division of Fuel Chemistry, 157th Meeting, ACS, Minneapolis, Minn., April 13-18, 1969.
EPOXIDATION OF FISH OIL Kinetic and Optimization Model J A I M E W l S N l A K
A N D
E D U A R D O N A V A R R E T E
Department of Chemical Engineering, Gnicersidad Catolica de Chile, Santiago, Chile Anchovy oil of iodine value 188.8 was epoxidized in situ, with preformed peracetic acid, or using a mixed strategy, to determine a kinetic model for the reaction arid the optimization of double bond conversion to oxirane rings. In the range of the operating variables, epoxidation and ring opening may be described by a pseudo-first-order reaction, and use of a mixed strategy allows a 92.2% conversion of the double bonds with a final oxirane number of 8.5 and iodine number of 19.4. Optimal conditions correspond to an in situ process a t 70' C. w i t h partially preformed peracetic acid and fast addition of the oil, using 6.1 6 moles of 41 weight Yo hydrogen peroxide and 0.5 mole of acetic acid, per mole of ethylenic unsaturation, in the presence of 10 weight O/O dry basis of resin catalyst Dowex 5OW-l2X, 50/100-mesh. Thermal stability of the epoxidized oil compares favorably w i t h that of commercial PVC plasticizers.
EPOXY plasticizers for polyvinyl
resins are usually prepared from vegetable oils that are cheap and have a low iodine value. T o produce plasticizers with an oxirane number of 9 or higher, it is necessary to improve the unsaturation of the oil, using processes such as the Solexol, which consists in a liquid-liquid segregation with propane or other solvents. The iodine value may thus be raised to about 200 with a corresponding increase in cost of the raw material. I t would seem appropriate to explore the possibility of using other cheap raw oils, like fish
oils, that have a relatively high iodine value (180 t o 220). Few references are available in the literature as to the possibility of using this type of material (Greenspan and Gall, 1953, 1956, 1957). Previous work (Wisniak et al., 1964) showed that standard procedures may be used to epoxidize fish oils and attain oxirane numbers of about 7.8. Unfortunately, the final product presented a high residual iodine value and did not have the proper thermal resistance to be used a t high temperatures without decomposition. Ind. Eng. Chem. Prod. Res. Develop., Vol. 9,No. 1, March 1970
33
