I
E. W. LANG and J. C. LACEY, Jr. Southern Research Institute, Birmingham 5, Ala.
Properties of Tar from
...
l o w Temperature Carbonization of America Seam Coal Study these proposals carefully-they may grow into an entirely new industry
IN
THE future, large quantities of lowcost tar may be available from recently developed processes for fluidized carbonization of coal. T h e aim of such carbonization processes is to provide a lower cost fuel char for the generation of power by the sale of the tar. However, finding outlets for this tar is difficult because its physical and chemical properties differ from coke-oven tar and also because the quantities which may be prolarge steam duced are so large-one generating power plant might produce 50,000,000 gallons per year. Some work has been done in determining properties, composition (7, 9, l o ) , and uses (1-6, 8, 72, 15, 16) of this tar, but little technical information is available from these programs. T h e work described here was done to evaluate tar produced from Alabama bituminous coal and to investigate some methods for utilization. I t now appears that the conversion of tar into chemical products on a commercial scale is a matter of 5 to 10 years in the future. When petroleum and residual fuel prices increase to approximately 10 cents per gallon at inland coalproducing areas, operation of commercial carbonization plants should be profitable through the sale of tar as a liquid fuel. Once low-temperature carbonization is under way through sale of tar as fuel, then commercial processing of the tar to chemicals and higher-priced products will follow.
Experimental Program
This program on tar utilization was undertaken to provide a sounder basis for evaluating the potential of low-temperature carbonization for the sponsoring power utility. A prior program (73)for this power utility had demonstrated the feasibility of carbonizing agglomerating coals in a fluidized-bed process. Cost estimates indicated that carbonization would be economic if the tar could be sold for 8 to 10 cents per gallon. Disposal of the tar as a heavy liquid fuel would not always bring this required price. Therefore, for profitable disposal of the tar, some premium values of the tar must be realized. An investigatory program was car-
ried out to evaluate in a preliminary manner various means of utilizing or converting the tar from the Alabama coals, to determine approximate yields of salable products, to reveal avenues of research that should be followed, and to establish a n approximate price for the tar in terms of processing costs and prices of marketable products. T h e program was not intended to develop detailed and optimum procedures for each type of processing that was considered. Thus, the results from this program are not precise, and the operating conditions and yields are probably not the optimum values that further research will provide. Tar Composition. Lowtemperature
tar has undergone less thermal cracking than coke-oven tar and therefore is less aromatic. I t may be regarded as intermediate between coke-oven tar and crude petroleum, but very few individual chemical compounds are present in amounts sufficient for economical recovery. T h e tar used in this program was produced by carbonizing America Seam coal, a high-volatile bituminous coal from A41abama,at 950' F. in a continuous fluidized-bed carbonization pilot plant (13). This tar is similar in physical and chemical properties to tars produced in other low-temperature carbonization processes from a variety of coals (see table).
Properties of Tar from Low-Temperature Carbonization of America Seam Coal Speciflc gravity at 2 5 ' / 2 5 ' C. = 1.10 Distillation Yields (Hempel Method) (11) Compobition of Low Hoilerc (to 235' C.) Temp.. C . 1Vt. 9o T'ol. yo 28.5 To 170 1.4 Tar acids 2.2 170-200 3.3 Tar bases 69.3 Neutral oils 200-235 7.6 235-270 270-300 300-360
Pitch Loss
7.3 7.5 14.3 51.8 6.8
Coml>ositionof Distillate VOl.
Tar acids Tar bases Neutral oil
yo
28.5 2.9 68.6
Distillation of T a r .\rid. Temp.,' C. Wt. % 170-200 (phenol) 3.1 200--210(cresols) 13.9 210--235(xylenols) 22.0 235-300 Above 300 and loss
20.8 40.2
Composition of Dibtillate Nciitral Oil Vol. yo Olefins 8.1 47.3 Aromatics Saturated hydrocarbons 44.6
C~ompositioiiof Pieutral Oils in Low Boilers Vol. % Olefins
Aromatics Saturated hydrocarbons
11.2 38.2 50.6
Composition of Saturated Hydrocarbons 1'01. % Naphthenes 41 Paraffins 59 Distillatinn nf Heavy Distillate (235-360' C.) Wt. yo To 235' C. 6 235-270 270-360
14 48 32
Residue
Composition of Heavy D i d l a t e Vol. % 35.7 Tar acids 3.4 Tar bases 61 .O Neutral oil Cornpobitioii of h-eutral Oil ill Heavy Distillate Vol. y* Aromatics 51.6 19.2 Olefins Saturated hydrocarbons 29.2
VOL. 52, NO. 2
FEBRUARY 1960
137
+
20 000000 000000
3
% 3 0
m
o
M O
I,
3
6
m *
x
-m 0 m
a m
3 .4
4
cu
. Y , . I
u m
3
b 000 0000 00 ' 0 '
0008 0'
rw(wwc o oo 4
ri I
8 u Q,
Y
0)
3
a E 0
V
Ll 0 c
m
m Q,
u
0
cta :a: aJ
0
k.. c*
6,
z
U
138
INDUSTRIAL
AND ENGINEERING CHEMISTRY
CHEMICALS FROM COAL C A R B O N I Z A T I O N T h e distillation yields are similar to coke-oven tar and also to some heavy high-sulfur crude oils. Light Distillate. T h e light distillate fraction boiling below 235" C. has a much greater value than the other fractions because approximately 30% of it consists of phenol, cresols, and cresylic acid. Aromatics (4070 of the remaining neutral oil) can be recovered, but rhe oil would be more valuable if the aromatic content of the oil were increased. T o study the possibility of converting the large amounts of naphthenes in the light distillate to aromatics, two runs under reforming conditions were carried out in a bench-scale continuous pressure reactor containing 550 cc. of '/*-inch molybdena-alumina catalyst. Space velocity was 0.4 volume,'volume/ hour; temperature was 525' C.; and hydrogen pressure was 370 psi. T h e charge was neutral oil from the light distillate. T h e aromatics content of the oil was increased from 38 to 6570 a t a yield of 717' (volume). Nitrogen and sulfur coinpounds were practically eliminated. Heavy Distillate. Using the heavy distillate, boiling between 235 ' and 360 C.. four batchwise hydrogenation runs were made with maximum hydrogen pressures of 3000 to 4000 p.s.i., temperatures of 450' to 500' C., and a molybdena-alumina catalyst. After a reaction time of 1.5 hours, the yield of oil was 93% (volume), 45y0 of the product boiled below 235' C., and the product contained 5370 aromatics. T h e consumption of hydrogen was 5.570 of the weight of the tar. These runs indicated that 475' C. was about the optimum temperature, and yields were much lower with longer holding times and higher temperatures. Continuous runs a t 480' C. were made in a bench-scale unit which had a reactor catalyst volume of 550 cc. T h e hydrogen pressure was 3000 p s i . ; space velocitv was 0.4 to 0.8; and the catalyst was '8-inch pellets of molybdena alumina. Under these conditions, little of the aromatic compounds was hydrogenated, and the desired degree of cracking to the low-boiling compounds was accomplished; tar acids were practically eliminated. The results from the continuous runs were quite similar to those from the batchwise runs. On the basis of five runs, the best yields a t these conditions wcre approximately 92y0 (volume), and the amount of the product boiling below 235' C. was 50%. T h e product contained 40% aromatics. T h e hydrocracked oils boiling below 235' C. from these runs still contained considerable amounts of naphthenes, and their conversion to aromatics via reforming was studied in a continuous run.
T h e equipment and conditions were the same as used for reforming of the lowboilers distilled from the tar. An oil yield of 7470 (volume) was obtained. T h e product contained 80% aromatics as compared to 447, aromatics in the charge stock. Neutral Heavy Distillate. I t is probable that in commercial processing of the distillate oils, the t a r acids will be extracted for separate processing before the oils are catalytically treated because such treatments destroy the tar acids. A number of continuous runs was made with this neutral heavy distillate to determine the effect of thermal cracking a t atmospheric pressure with and without a catalyst and the effect of low-pressure reforming-cracking conditions. The neutral heavy distillate used for these runs was the fraction of the total tar boiling from 235' to 360' C. after the tar acids were extracted with caustic washing. Thermal cracking runs on the neutral distillate made without a catalyst a t 600' C. and a t atmospheric pressure reresulted in yields of approximately 80% (volume). Only about 17y0 of the product boiled below 235' C. Another series of runs was made using molybdenaalumina catalyst and hydrogen a t atmospheric pressure and a space velocity of 1 .O. At a temperature of 640' C., the yield was 55970, and the product contained 777' aromatics and 227, oils boiling below 235' C. .4t a lower catalyst temperature of 520' C. the yield was 87%, but the product contained only 167~ oils boiling below 235' C. These runs indicated that satisfactory yields of cracked products can be obtained only when a catalyst and sufficient hydrogen pressures are used to prevent cracking from proceeding too far. Hydrocracking-reforming of neutral heavy distillate was made a t fairly high temperatures to give sufficient cracking and dehydrogenation and a t moderate hydrogen pressures to prevent overcracking of the compounds so that good liquid yields would be obtained. T h e selection of conditions was based partially upon a series of batch reactions with heavy distillate. With the use of a molybdena-alumina catalyst a t about 480' C., hydrogen pressure of 750 p.s.i., and space velocity of 1.0, yields were 75 to 91% (volume). T h e product contained 73 to 80% aromatics and 40% boiled below 235' C. With a temperature of 550' C., hydrogen pressure of 500 p.s.i., and a space velocity of 0.4, the yield was 74% (volume), the product contained ?9% aromatics, and 4670 of the product boiled below 235' C. I n a commercial operation, the oils boiling above 235" C. would probably be recycled to the reaction so that only lowboilers would be produced. O n e run
was made to determine the effect of recycling the high-boiling oil. T h e yield of product was 66% (volume). T h e product contained 93% aromatics, and 2670 boiled below 235' C. T h e combined reformate boiling below 235' C. from the primary and recycle runs was carefully fractionated and the aromatics in each fraction determined. This analysis indicated that the combined chemical oil contained 4% benzene, 8% toluene, lOy0xylenes, 16% naphthalene, 30% substituted benzenes, and 32% saturated hydrocarbons. Tar Acids, Approximately 4070 of the tar acids from the total distillate consists of phenol, cresols, and xylenols. About 60Y0 consists of high-boiling, highly substituted alkyl phenols for which there is little demand a t the present. T h e research program has shown that the tar-acid fraction boiling from 170 to 235 O C. (including xylenols) will form hard thermosetting resins with formaldehyde. T h e removal of small amounts of impurities from individual tar acids appears to be a major problem. O n e means cf utilizing the tar acids is to convert by dealkylation-hydrocracking the higher-boiling ones, which have little value, to the simpler, more valuable lowboiling phrnols. T a r acids boiling from 235' to 360° C. were hydrocracked in a batch reactor a t 420' C. in the presence of molybdena-alumina catalyst and 10% added water, and a t a hydrogen pressure of 2600 p.s.i. T h e yield was 92% (volume). T h e product contained 82% tar acids and 237, oils boiling below 235' C. T a r acids boiling in the intermediate range of 235' to 300' C. were hydrocracked a t 460' C. and 2100 p.s.i. hydrogen pressure to give a yield of 90% (volume). T h e product contained 507c tar acids, and 5370 boiled below 235' C. Hydrocracking of the heavy tar acich boiling a t 300' to 360' C. resulted in a yield of 84y0 (volume). T h e product contained 50% tar acids, and only 23% of the product boiled below 235' C. Pitch. Approximately one half of the tar consists of hard pitch. 'Therefore, no plan for conversion of the tar will be successful unless attractive markets or methods of conversion for the pitch are found. T h e pitch might be sold in the large-volume markets of road asphalt or pitch, roofing pitch, electrode binder pitch, or for a premium fuel in openhearth furnaces of steel mills. T h e pitch might also be converted to pitch coke or hydrogenated to give more of the lowboiling chemical oil. Preliminary investigations were made for adapting the pitch to these various outlets. Additional work is needed on making the pitch more stable to weather if it is considered for uses such as road and roofing pitch. Electrode binder pitch must contain a t VOL. 52, NO. 2
FEBRUARY 1960
139
least 20% of so-called “beta resin,” which is that portion soluble in quinoline but insoluble in benzene. A series of experiments demonstrated that the beta resin content of the pitch could be increased to the range of 30 to 50% in yields greater than 95% by air-blowing a t 250’ C. No catalyst was used in any of the runs. Coking of the pitch was studied in five batchwise retort runs made a t 700’ C . to determine the yields of pitch coke and oils. T h e average yield of coke was 51% and that of tar was 31y0, with the remainder being water and gas. T h e coke had a volatile content of 5%. T h e tar from the coking of pitch was very viscous and contained only 19% of oils distillable to 360’ C. Tar Hydrogenation. A limited number of batch hydrogenations was carried out using the total tar which had been topped to 235’ C. Such hydrogenation of the total heavy tar might be considered for commercial utilization of the tar if previously mentioned uses of the tar d o not provide sufficiently attractive outlets. Low-temperature tar is much more reactive than coke-oven tar, and has been hydrogenated to low boilers in yields of approximately lOOyo (volume) in other research programs (72). T h e runs made in the present program were carried out a t temperatures of 450’ and 470’ C. and a t maximum hydrogen pressures of about 3000 p.s.i. Yields of 90 and 98% (volume) were obtained, and 20 to 30y0 of the product boiled below 235’ C. T h e conversion to low boilers was not so great as desired, and further study is needed.
Methods for Processing Tar to Marketable Products
Potentially, this tar has a number of uses, which range from crude uses such as fuel and road tars to complete conversion to low-boiling aromatics. Several schemes for partial or total conversion were considered on the basis of the results from the preliminary experimental program and on published data. T w o processing methods were evaluated based on a plant processing 41,600,000 gallons of tar per year produced by fluidized carbonization of 6000 tons of America Seam coal per day. T a r was charged to the tar processing plant a t 8.6 cents per gallon because this price would give a net return of 10% after income taxes on a carbonization plant located in Alabama. These costs were based partly on a n engineering cost study of fluidized carbonization by United Engineers and Constructors, Inc. ( 7 4 ) . T h e first of the two proposed plans for tar processing involved using a portion of the tar without chemical conversion and converting the remainder by hydro-
140
cracking. T h e low-boiling oils are removed and refined to tar acids and a neutral oil. T h e pitch and nonaromatics from the heavy distillate are sold as road binder. T h e aromatic portion of the heavy distillate is hydrocracked to lowboiling oils. T h e refined products consist of low-boiling aromatics, cresylic acid, a n d gasoline blending stock. T h e total estimated plant cost is $6,470,000, of which $2,628,000 is for the hydrocracking unit and a n additional $1,310,000 is allowed for working capital. T h e processing costs include amortization, direct and indirect production costs, raw material costs, and sales and administrative costs. T h e estimated net income after income taxes is $979,000, which gives a net yearly return on the investment of 12.676, or a payout time of four years. T h e second plan was for the complete conversion to low-boiling products. This plan was proposed for the eventuality that the pitch and heavy distillate cannot be sold in existing markets or would bring only fuel prices. This method of utilization calls for removing the tar acids boiling below 300’ C. and hydrocracking the rest of the tar to oils boiling below 235’ C. T h e detailed steps in this method of refining are indicated in the flow sheet. Oils boiling below 300’ C. are topped from the total tar and are separated into tar acids and neutral oils by solvent extraction. T h e neutral oils are combined with the heavy tar as feedstock to the hydrocracking unit. T h e tar acids from solvent extraction are distilled to give low-boiling (licacid which are sold. Hydrocracking of the heavy tar and the neutral oils from the tar-acid separation step is carried out a t 3000 p.s.i. and 480’ C. in the presence of molybdena-alumina catalyst. T h e product is topped, and the heavy end is recycled to the hydrocracking step. T h e lowboiling fraction boiling below 235’ C. is catalytically reformed to increase the aromatics content to SOY0. Solvent extraction of the aromatics and subsequent fractional distillation gives benzene, toluene, xylene, naphthalene, and aromatic solvent. T h e nonaromatic portion of the reformate is sold as gasoline blending stock. Over-all yield of finished products is 68 gallons per 100 gallons of crude tar. T h e estimated cost of the plant is $13,152,000, including working capital. T h e total yearly costs for processing the tar are estimated a t $3,733,000. T h e net income after income taxes for this plan is $912,000, which provides a return on the investment of 6.9yc.
INDUSTRIAL AND ENGINEERING CHEMISTRY
Present Outlook
T h e return on the investment of either of these plans is lower than that expected by most chemical companies for processes that will involve heavy expenses for development work. However, several factors may improve the economics of processing this tar. Furthrr research may result in both higher yields of the valuable products and simpler processes. Also, long-term higher costs of competitive raw materials and finished products will better the economics of tar processing, and further work on coal carbonization processes should give higher yields of tar and lower costs of carbonization. O n e the other hand, the general recession in 1958 caused prices to soften for the aromatic chemicals and certain fuel fractions for which some of the tar would be used. I t may be several years before these losses may be overcome by rising prices. T h e increasingly large amount of imported petroleum of the past few years has softened the price of fuels and petroleum products with which some of the tar products must compete. Acknowledgment
T h e authors thank the iilabama Power Co. and its affiliate, Southern Services, Inc., sponsors of the work, for permission to publish the results of the program, and R. E. Lacey and H. G . Smith for their assistance. literature Cited (1) Chem. Eng. 6 5 , No. 2 , 63 (1958). (2) Ibid., 64, No. 7 , 228 (1957). (3) Chem. Ene. News 36, 55 (June 30, ~, 1958). (4) Ibid., p. 27 (May 26, 1958). ( 5 ) Ibid., 34, 1318 (1956). ( 6 ) Clough, H., IND. ENG. CHEM.49, 673 (1957). (7) Gomez, M . , Goodman, J. B., Parry, V. F., U. S. Bur. Mines, Rept. Invest. 5302 (February 1957). (8) IND.ENG. CHEM. 49, Pt. I, 3 2 A (September 1957). (9) Karr, C . , Jr., Brown, P. M., Estep, P. A,. HumDhrev. , , G. L., Anal. Chem. 30, 1413 ’(1958)A. (10) Karr, C., Jr., Chang, T . C. L., J . Inst. Fuel 31, 522 (1958). (11) Kester, E. B., Pohle, W. D., Rockenbach, L. P., Bur. Mines, Rept. Invest. 3171 (1932). (12) King, J. G., “The HydrogenationCracking of Tars,” in “Science of Petroleum.” vol. 111. D. 2157, Oxford Univ. Press; London, l j 3 8 . 3) Lang, E. W., Smith, H. G., Bordenca, C., IND.ENG.CHEM.‘?~, 355 (1957). 4) Minet, R. G., Economics of the Continuous Fludization Carbonization Process for Bituminous Coal,” American Gas Association Meeting, Bar Harbor, Fla., May 1957. 5) Petrol. Week 5 , 40 (Oct. 25, 1957). 6 ) Pound, G. S., Coke and Gas 1952, 355 (October). I
RECEIVED for review June 1, 1959 ACCEPTED September 30, 1959 Division of Gas and Fuel Chemistry, Symposium on Tars, Pitches, and Asphalts, 135th Meeting, ACS, Boston, Mass., April 1959.
