Ind. Eng. Chem. Process Des. Dev. 1984, 23,637-641 McNell, D. “High-TemperatureCoal Tar” In "Chemistry of Coal Utlilzation”, Elliott, M. A., Ed.; W k y : New Yo&, 1981; p 1003. Panson, A. G.; Kovach, J. J. “AnalytlcaiData-F’roducerlFuli Flow Cleanup System,Run #Q7”,METC, IR No. 1080, July 24, 1981. ROSS, c. s.; tienblcks, s. B.us.w .SWV. R O ~ S S . m w IO~L), 2058. Sawada, K. M.S. Thesis, Department of Chemlcal Englnwlng, M.I.T., 1982. Sircar, S.; Kumar, R. ACS Svmp. Ser. 1083, 223, 195-212. sokmon. P. R.; Hamblen. D. 0. EPRI AP-2603, Project 1854-8. Flnal Report, July 1983. Swberg, E. M. Sc.D. Thesls, Department of Chmical Engineering, M.I.T., 1977. Swberg, E. M.; Peters, W. A.; Howard, J. B. 17th International Symposlum on Combustion, 1978.
637
Wilson, S. T.; Lok, B. M.; Messlna, C. A,; Cannan, T. R.; Fianlgen, E. M. Abstracts, 184th National Meetlng of the Amerlcan Chembl Society, Kansas City, MO, INOR-008, 1982.
Received for review March 17, 1983 Revised manuscript received December 1, 1983 Accepted December 21, 1983 The research work was undertaken with a financial report of Morgantown Energy Technology Center, U.S.Department of Energy (Contract No. DE-AC21-80M14385).
Influence of Physical and Chemical Parameters on Wood Pyrolysis Ollvler Beaumontt and Yvan Schwob’ E W Natbnale SUperiewe des Mines de Paris, Centre Reacteurs et Recessus, Equip de Recherches Ass& 768, Laboratoke de pvrolve des Blomesses. 60, Bd Saint-Michei, 75272 Paris Cedex 06, France
au CNRS N o
An original pyrolysis untt has been developed to study the influence of pyrolysis parameters: temperature, heating rate, wood particle size and moisture, gaseous environment, and catalyst impregnation. The effects of these parameters on char yield, dl and gas yield, and composttion are presented and interpreted to assist the industrial application of wood pyrolysis technology.
Introduction The search for alternative energy supplies has renewed interest in the utilization of agricultural and forest byproducts as substitutes for conventional sources of energy. In this area, wood pyrolysis is of special interest as it produces pyrolytic oil as well as energy. Pyrolytic oil differs from natural oil in that valuable chemicals are present in a complex mixture as they are not destroyed by a coarse thermal process. At present, pyrolytic oil, the major product of wood pyrolysis, is seldom processed for chemical recovery. Pyrolytic oil is a very complex mixture; it differs from natural oil that basically contain hydrocarbons. In addition, pyrolytic oil is unstable, leading to considerable difficulties in the separation process. The scope of this study is the investigation of the pyrolysis phenomena in order to control it better. Experimental equipment was developed for testing physical and chemical parameters such as temperature, particle size, gaseous product, extraction, wood moisture, and the influence of catalyst. Experimental Section Beech wood (Fagussylvatica) was selected for the study as the typical European hardwood. The particle size appears to be a determinant parameter; thus the wood sawdust obtained from the original piece of sound wood was sieved to three different particle sizes: fine, from 0.050 to 0.125 mm (referred to as “F”),medium, from 0.125 to 0.25 m m (referred to as “M”), and course, from 0.25 to 0.5 mm (referred to as “G”). The chemical composition of beech wood obtained from standard analytical methods (PETROFF and DOAT, 1978) is presented in Table I. Wood moisture is also a relevant parameter. It is determined by oven drying at 105 “C to constant weight. ‘Centre de Recherche Elf-Solaize, Section Proc&l6s-Induatrie, 69360 Saint-Symphorien d‘Ozon, France.
Samples of different moisture were prepared by moisturing and/or drying. Temperature is the most determinant parameter; not only the final temperature of the treatment but also the heating rate is important. An experimental device was developed that allows testing of both parameters. The experimental setup allows pyrolysis of wood particles in a gaseous sweeping stream (Figure 1). The gaseous stream rapidly sweeps the pyrolysis products out of the furnace where they are condensed and cooled. This precaution avoids as much as possible the secondary degradation of the volatile products. An “extractive”wood pyrolysis is thus carried out. The extractive gas flow can be a permanent gas, such as helium or nitrogen or an superheated solvent vapor. Methanol, 2-methylpropanol, and ethylglycol (C2H60CH2CH20H)were used because of their efficiency as solvents of all the constituents of pyrolytic oil. The solvent is placed in a pressurized tank and flows out through an expansion valve that ensures a steady flow. The solvent is vaporized and superheated in the packing at the bottom of the furnace. When permanent gas is used for sweeping, a supplementary tubing with peristaltic pumping equipment allows one to recirculate part of the gas to obtain high specific flow rate in the furnace without excessive dilution of the pyrolytic products in the sweeping gas. The recirculating gas is taken after recovery of pyrolysis products. The temperature of the furnace is controlled by a programmed regulator monitored by a chromel-alumel thermocouple placed near the heating coils of the furnace. Another thermocouple of the same type placed in the wood bed measures the effective temperature of the wood. The temperature gradient in the furnace was found to be homogeneous radially but there was a small vertical dispersion throughout the volume of the wood that could not be reduced to less than 4”. The condensed pyrolytic oil is collected for analysis in a cold trap and the gaseous products are collected in a tight
Q79543Q5fa4f7 ~23-a637$a1.5afQ0 1984 American Chemical Society
638 Ind. Eng. Chem. Process Des. Dev., Vol. 23, No. 4, 1984 Table I. Chemical Composition of Beech Woodu constituents % ethanol-benzene (1/1)extract on extracted sawdust water extract NaOH extract pentosan
a
0.9 1.3 18.3 26.8
cellulose
40.1
lignin ash silica
23.2 0.5 0.001
analytical method used 7 h in a Soxhlet apparatus 7 h refluxing 7 h refluxing hydrochloric acid (13.5%) reactant, distillation, of furaldehyde, and titration with bromide-bromate 3 treatments with nitroalcoholic mixture and correction of the remaining ash and pentosan sulfuric acid method constant weight at 425 "C in air nitroperchlorhydric reactant
Proximate analysis: fixed carbon, 24.5%; volatiles, 75.0%; ash, 0.5%;moisture, given for each experiment.
r-
I
Figure 2. Gas analysis setup for CO, COP,N z , Os, H P , and light hydrocarbons determination: (1) chromatographer oven; (2) sampling valve; (3) sampling loop; (4) carrier gas (N2); (5)first column; (6) second column; (7) catharometer.
Figure 1. Experimental setting for sawdust pyrolysis: (1)vibratory device; (2) feedstock (wood); (3) pyrolysis chamber; (4) electric furnace; (5)condenser; (6)cold traps; (7) tight gas tank; (8) sweeping gas stream; (9) pressurized solvent tank; (10) gas volume measure; (11) peristaltic pumps; (12)packing materid; (13) thermocouple.
tank where they are homogenized before analysis. The volume of gas produced is then measured by a gasmeter. Two different experimental procedures are used to study the effect of temperature and heating rate. (a) Flash Pyrolysis. The furnace is preheated at the desired temperature and the sweeping gas stream is established. The wood powder is then introduced by a vibratory device into the furnace. The rate of introduction is adjusted so that the furnace temperature is not disturbed: 10 g of wood is introduced during each experiment at the rate of 0.5 g/min. Afterwards, the temperature is maintained for completion of the processing. (It was found that the char residue underwent no additional transformation after 30 min.) The furnace is then rapidly cooled. The char is collected and weighed. This procedure approximates closely isothermal reaction. (b) Slow Pyrolysis. Wood powder is placed in the cold furnace and the sweeping gas stream (10 g) is established. The furnace is theh heated a t a definite rate. Gas Analysis Pyrolytic gas was analyzed by gas solid chromatography for carbon monoxide, carbon dioxide, nitrogen, oxygen and light hydrocarbon determination. A device made of two analytical columns in series shown in Figure 2 was used with flow of gas through each side of the catharometric detector of a chromatograph Girdel 30 (Jecko and Reynaud, 1967). The first column C1 (2 m, l/g in. stainless steel) is packed with Porapak R and separates carbon dioxide, hydrocarbons, and a mixture of air and CO + methane. The second column (2 m, in. stainlea steel) packed with a 5 A molecular sieve, separates
oxygen, nitrogen, methane, and carbon monoxide; the other components are adsorbed irreversibly. The analytic output is processed by an LTT-ICAP integrator and recorded on a Sefram Servotrace recorder. Quantitative gas analysis is performed by internal normalization. Individual response factor of each gas is determined on a quantity of pure gas precisely injected with a computer-monitored HAMILTON syringe. Pyrolytic Oil Analysis The pyrolytic oil is analyzed by gas-liquid chromatography, on a Girdel 3000 chromatograph, with catharometric in. stainless steel column packed detector using a 2 m, with Porapack Q. The same integrator and recorder are used. Only the most important components are isolated by this method but it is enough to test the influence of physical and chemical parameters. Further study of more complete analysis of oil is in progress. The quantitative determination of the components is performed by internal standard (ethyl acetate). The accuracy is satisfactory, except for methanol that is not sufficiently separated from water and formic acid which is difficult to analyze by GLC because of its polarity and tendency to dimerize in the gas phase. Results and Discussion After weighing, the char, pyrolytic oil, and gas, the total weight obtained was typically 97 to 100.5% of the reacted wood. The intluence of temperature was studied in a series of flash pyrolyses of sawdust at 8% moisture with nitrogen as the sweeping gas. The evolution of yields of char, pyrolytic oil, and gas with the temperature are presented in Figure 3. Four areas can be established (1)drying area under 220 "C; (2) roasting area from 220 to 330 "C; the solid residue is predominant: it is a roasted wood of dark brown color; (3) pyrolysis area from 330 to 450 O C : a true char is ob-
Ind. Eng. Chem. Process Des. Dev., Vol. 23, No. 4, 1984 630 Table 11. Pyrolytic Oil Composition a slow pyrolysis flash pyrolysis
a 100 0
temp, "C
300
3 50
400
450
500
water methanol ethanal acetone formic acid acetic acid propionic acid 1-hydroxypropanone 1-hydroxy-2-butanone 2-furaldehyde furf urylic alcohol
9.1 0.68 0 0.1 1.21 5.79 0.18 1.30 0.51 0.76 0.36
13.1 0.86 0.03 0.08 0.43 6.91 0.14 2.55 1.66 0.69 0.43
14.9
18.6 1.02 0.17
16.3 0.97 0.45
7.84 0.45 3.57 2.49 1.04 0.24
1.88 0.87 0.44
17.3 1.54 0.17 0.25 6.32 0.20 2.66 1.32 0.79 0.63
6.87 0.55 2.51 2.00 1.07
