Ind. Eng. Chem. Process Des. Dev. 1982, 21, 530-532
530
I6O
140 120
I 1
COhlBiNEU
/
6
l\ICFIEASE Clr'ER EASE CAS6
o2
N E ~ . R C-IED A
a
BLAS-
Figure 4. Effect o f oxygen enrichment on gasifier output.
be composed of elemental carbon. Table I1 compares the predicted product gas compositions with those reported by Batchelder et al. for two cases involving gasification with air (21% 0,) and gasification with a 50.7% oxygen blast. These results established the basis for the oxygen-enriched cases, which were examined in increments of 5% oxygen. The incremental steam required for additional cooling at higher O2levels was predicted on the basis of volume-for-volume substitution of nitrogen, as suggested by Batchelder et al. Figure 3 shows the resulting heating values, HHV and LHV, for the enrichment cairn as predicted by the model. For example, a 10% enrichment level yielded a 16% predicted improvement in heating value. The production incentive for an oxygen-enriched blast in anthracite gasification is based not only on the incremental improvement in heating value, but also on additional capacity, which results in a significant increase in total Btu production. These effects are shown in Figure 4 (lime B) as a function of oxygen enrichment over the base
case using air (21% OJ. Capacity improvements were calculated from the oxygen-carbon ratio extrapolated from the Kerpely producer data. When the effects of higher Btu content and increased capacity were compounded, the result was a combined production factor, also shown in Figure 4 (line C). The data indicated that at a 10% enrichment level, the total production improvement would be approximately 100%. It is interesting to compare the two-zone model advanced here with a similar "kinetics-free" (KF) model described by Denn et al. (1979). In the KF model, the CO:C02 ratio in the gas leaving the combustion zohe is arbitrarily fixed, and the corresponding combustion zone temperature is computed from the adiabatic temperature rise in the reactor. The "combustion zone" temperature for fixed-bed anthracite gasification predicted by both models lies with 2750-2950 OF. These temperatures are somewhat higher than expected when ash fusion limitations are considered. Denn et al. have calculated that variation from 0 to 100% of the CO fractional conversion used in the KF model causes a variation of only 300 OF in the computed maximum temperature. Conversely, investigations with the two-zone model developed here indicate that the "combustion zone" temperature is relatively sensitive to the fractional heat leaks (typically 3% of the feed HHV) normally ascribed to a commercial gas producer operation. Conclusion A model has been developed to simulate fixed-bed gasification of anthracite and similar feedstocks. Results compare favorably with published data which show the effects of oxygen-enrichment of the air blast. The model predicts improved heating value of the product gas as the oxygen concentration of the air/steam blast is increased. Additionally, potential capacity increases are identified. Literature Cited Batchelder, H. R.; Dressier, R. G.; Tenny, R. F.; Kruger, R. E ; Segur, R. D Am. Gas Assoc. R o c . 1950, 340. Denn, M.; Yu, W,; Wei, J. I d . Eng. Chem. Fundam. 1979, 18, 286. Jennings, E. H. "Environmental Engineering"; International Textbook Co : Scranton, PA, 1970
Received for review May 30, 1980 Revised manuscript received February 1, 1982 Accepted M a r c h 8, 1982
COMMUNICATIONS Sorption of Sulfur Dloxide by Waste Calcareous Materials The conversions of calcareous muds from a sugar mill and acetylene production plant ?ere studied in terms of exposure time and temperature. There is an essential difference in the course of the sulfation reaction of these materials and that of common limestones. While conversions of common limestones are, in general, low, the calcareous muds retain a considerable reactivii until the complete conversion of calcium oxide to sulfate is attained.
Introduction It is a well-established fact that particles of limestone exposed at high temperature to the flue gas containing sulfur dioxide react only partially. This poses a major 0196-4305/82/1121-0530$01.25/0
obstacle to the practical feasibility of dry limestone processes for removal of sulfur dioxide from combustion gas (Krijger et al., 1981). In our previous work we investigated structural changes 0 1982 American Chemical Society
Ind. Eng. Chem. Process Des. Dev., Vol. 21, No. 3, 1982
531
Table I. Analysis of Calcined Calcareous Materials component (% by weight)
material
CaO
MgO
A1203
Fez03
wt loss o n ignition at 850 " C
carbide mud sugar mill mud
82.8 79.5
1.26 1.78
3.30 1.43
0.861 0.566
0.450 0.454
of reacting limestone particles (Hartman and Coughlin, 1974), developed a mathematical model for the sulfation reaction (Hartman and Coughlin, 1976),and studied basic factors limiting the sorptive utilization of limestone (Hartman et al., 1978). Our experimental findings and conclusions as well as those of other authors (Potter, 1969; Falkenberry and Slack, 1969; Borgwardt and Harvey, 1972; Dogu, 1981; Bhatia and Perlmutter, 1981) clearly suggest that both the rate of reaction and the attainable conversions are closely related to the porosity and grain size of limestone particles. This brief study was inspired by an effort to find some alternative sorbents for sulfur dioxide. The experience shows that a material well suited for the sorption of sulfur dioxide should be fine-grained and porous. Unfortunately, such minerals are not common in nature. The dense, medium- to coarse-grained limestones and dolomites are usually available. For their rather amorphous structure the calcareous muds formed in the production of acetylene by the decomposition of calcium carbide and muds formed in the production of sugar have attracted our attention. The present work was undertaken to examine by experiment the sulfur dioxide reactivity of these waste materials at conditions of practical interest.
Materials and Methods The dry samples picked at plant dumps were mixed with some distilled water and extruded through a coarse sieve. The formed fragments were dried, crushed, and sieved. The fraction of particles within a size range 1.0-1.25 mm 1.12 mm) was investigated in this study. (Dh0th = mud materials are, in essence, mixtures of calcium carbonate and calcium hydroxide. The analyses of both materials provided similar results and are presented in Table I. Results Sulfation of the Carbide Mud. In the experiments the rate of sulfation was measured as weight gain. The sieved, dried particles were exposed to the simulated flue gas containing sulfur dioxide. A differential reactor with a thin, fixed layer of particles was used in this kinetic study. Further details on the apparatus and experiments can be found elsewhere (Hartman, 1974; Hartman and Coughlin, 1974). The results of our sulfation experiments are expressed in terms of conversion of calcium oxide to calcium sulfate. One should note that both materials also contain a small fraction of magnesium oxide which, to some extent, can also react with sulfur dioxide (Hartman and Pata, 1978). Since the amount of calcium oxide in the muds is about fifty times greater than the content of magnesium oxide, the reacted sulfur dioxide was related to the weight fraction of calcium oxide. The true conversions of calcium oxide are, therefore, slightly lower than those presented in this paper. In Figure 1are plotted the results of the sulfation experiments performed at 800 and 850 "C. I t can be seen that the rate of reaction is significantly affected by temperature. The conversions attained at 850 "C are higher
,
0.6r
0.L
1
0
k// I 15
i
I
L5
30
I
J
EO Cmin
Figure 1. Dependence of the conversion of calcium oxide in the carbide mud to sulfate on the exposure time and temperature: concentration of sulfur dioxide, 0.3% by volume; particle size, 1.12 mm; (0) temperature 800 O C ; ( 0 )temperature 850 "C.
01.1 0
I
I
20
LO
60 7,rnin
Figure 2. Dependence of the conversion of calcium oxide to sulfate in the sugar mill mud on the exposure time and temperature: concentration of sulfur dioxide, 0.3% by volume; particle size, 1.12 mm; (0)temperature 800 "C; ( 0 )temperature 850 "C.
than those measured at 800 "C. I t is interesting to note that the sulfation is very rapid, although quite large particles were employed (Dp= 1.12 mm). In the course of the first 5 min of exposure 44 and 55% of calcium oxide in such large particles was converted to sulfate. After 60 min high conversions of 86 and 99% were attained. The particles of the carbide mud employed in the work were not investigated more closely, but their considerable porosity and amorphous character were apparent. Sulfation of the Sugar Mill Mud. The experimental measurements were made in the same range and at the same conditions as those with the carbide mud. The results of the experiments are presented in Figure 2. In contrast to the preceding material, there is an apparent difference in the influence of temperature on the course of sulfation. The experimental data points obtained at 800 and 850 "C lie practically on a single curve. The weak effect of temperature is probably caused by compensation of a favorable influence of temperature on the chemical reaction itself and an adverse effect of temperature on the structure and texture of the formed calcium oxide. In the case of the sugar mill mud the rate of reaction is even higher than that at the carbide mud. It can be seen
532
Ind. Eng. Chem. Process Des. Dev. 1982,21, 532-535
03
06 ci2 9
min
Figure 3. Comparison of SO2 reactivities of the waste muds and different carbonate rocks: concentration of SOz, 0.3% by volume; (0) sugar mill mud; particle size, 1.12 mm; temperature, 800 O C ; ( 0 ) carbide mud; particle size, 1.12 mm; temperature, 850 “C; ( 0 ) limestone TI; particle size, 0.565 mm; temperature, 850 O C ; ( 0 )data of Borgwardt and Harvey (1972);on different types of carbonate rocks. The samples are numbered in the same way as in the original work. Concentration of SOz, 0.3% by volume; particle size, 1.3 mm; temperature, 980 OC.
from the data points plotted in Figure 2 that after 15-20 min of exposure, all calcium oxide was converted to sulfate. The numerical values of conversion slightly higher than unity are likely caused by the fact that magnesium oxide present in the material also takes part in the reaction. Comparison of the Waste Muds with Limestone. A fine-grained, high-grade limestone T I was chosen for the tests. This limestone is a common, dense carbonate rock used for the commercial production of lime. In accordance with previous experimental findings, conversions attained with the limestone were generally low. After 30 min of exposure about 15% of calcium oxide reacted with very slow conversion thereafter. The obtained results are plotted in Figure 3 and show a striking difference in sulfation of the mud particles and particles of limestone. The particles of the sugar mill mud and carbide mud react with sulfur dioxide at a rate much higher than the particles of limestone. Although the mud particles are two times larger, their sorption capacity to fix sulfur dioxide is about 5 times higher than the capacity of limestone.
Borgwardt and Harvey (1972) tested SO2 reactivity of a broad spectrum of naturally occurring rocks at similar conditions. Some of their results are also shown in Figure 3. The authors obtained the best results with marl which, in a similar way as the waste muds, can be completely converted to calcium sulfate. Like the sugar mill and carbide mud, the marl is also a porous, amorphous material. Conclusions The sugar mill and carbide muds are very efficient sorbents for sulfur dioxide. Even at advanced stages of the reaction the mud particles react rapidly with sulfur dioxide. In contrast to limestone, the complete conversion can be easily attained. The sorption capacity of both mud materials is about five times higher than the capacity of limestone. The difference in reactivity of the mud materials is not large. The sugar mill mud is somewhat more reactive. The rate of reaction of its particles is not practically affected by temperature in the range 800-850 O C . Nomenclature D, = mean particle size, mm t = temperature, O C X = conversion of calcium oxide to sulfate, mol of S03/mol of CaO Y = fraction of SO3 in sulfated sample, g of S03/g of calcine T = time of exposure of solid to gas, min Literature Cited Borgwardt, R. H.; Harvey, R. D. Envkon. Sci. Techmi. 1972, 6, 350. Bhatla, S. K.; Perlmutter, D. D. A I C E J . 1981, 27, 226.
Dogu, T. Chem. Eng. J . 1981, 27, 213. Fakenberry, H. L.; Slack, A. V. Own. Eng. Rug. 1989, 65(12), 62. Hartman, M. C d k t . Czech. Chem. C 0 ” u n . 1874, 39, 2374. Hartman, M.; Coughlin, R. W. I n d . Eng. Chem. Process D e s . Dev. 1974, 13, 248. Hartman, M.; Coughlin, R. W. A I C M J . 1978, 22,490. Hartmen, M.; Pata, J.; Coughlln, R. W. Ind. Eng. Chem. Process D e s . D e v . 1978, 1 7 , 411. Hartman. M.; Pata, J. Int. Chem. Eng. 1978, 18(4), 712. Kriiger, H.; Howlnd, HA?.; Schiylerl, K. Ger. Chem. Eng. 1981, 4 , 95. Potter, A. E. Am. Ceram. Soc. Bull. I96& 48, 855.
Institute of Chemical Process Fundamentals Czechoslovak Academy of Science 16502 Prague, Czechoslovakia
Miloslav Hartman* Karel Syoboda Jan Cermgk
Received for review July 14, 1981 Accepted March 29, 1982
Analytlcat Form of the Ponchon-Savarit Method for Systems wlth Straight Enthalpy-Composltbn Phase Lines An equation has been deriied which presents, in analytical form, the graphical construction of the Ponchon-Savarit diagram for systems with straight enthalpy-compositbn phase lines. This equation allows the number of theoretical plates to be computed and is convenient to use when a large number of stages is required. The minimum reflux ratio can also be calculated analytically.
Introduction The use of the Ponchon-Savarit method (Ponchon, 1921) to calculate the number of theoretical plates in a distillation column separating a binary mixture is a well-established procedure. The method permits graphical calculation of the number of stages required in the distillation column. In this paper, an equation has been derived which allows
the number of stages to be computed when the enthalpy-composition phase lines are straight. This equation is more general than Smoker’s equation (Smoker, 1938), which requires an additional assumption of equal heats of vaporization. An additional advantage is the ability to find the minimum reflux ratio analytically rather than by the trial and error method of Ponchon-Savarit. When the enthalpy-composition phase lines are linear, the Ponchon-Savarit method gives identical results ob-
0196-4305/82/1121-0532$01.25/00 1982 American Chemical Society
