Slaking of lime Johan B. Holmberg
Department of Chemical Engineering II, Lund Institute of Technology, P. O. Box 124, S-221 00 Lund, Sweden Quicklime is produced when limestone is heated above 900 ° C, a temperature at which limestone decomposes to carbon dioxid and quicklime quick lime (CaO). (CaO) . Quicklime exists exi sts in many different qualities. Limes of different quality have been characterised, wet slaked and dry slaked. The characterisation in combination with the wet slaking suggested that the relationship between the surface concentration of calcium and carbon was the best measure for the lime reactivity in the slaking process. The higher value for the relationship the more reactive was the lime. The wet slaking tests showed that the differences in reactivity decreases when temperature is raised or calcium chloride is added to the water. The dry slaking tests showed upon very small differences considering the slaking degrees for different conditions and different qualities of lime. However the tests showed on a relationship between water content in the reactor and the specific surface for the product, the higher quote the higher specific surface. Introduction
Quicklime is produced by heating limestone above 900 ° C, a temperature at which limestone decomposes to carbon dioxid and quicklime (CaO). The product quality is among other things dependent on this calcining process. Prolonged heating or heating at too high temperatures generate a less reactive lime. Another problem is how the calcining process is performed, what sort of furnace that is used. Besides how the calcining process is done, is the product quality governed by the amount of impurities in the lime. Quicklime is often used in many processes involving slaking, it is therefor of utmost importance knowing whether the quicklime will be slaked or not in the slaking process. To test this quicklimes of different qualities were investigated. The limes tested were a couple of Chinese limes, a Swedish lime and a lime from Poland. Characterisation of quicklime
The normal measurements when quicklime is characterised are the amount of free lime and total lime in the product,
the porosity of the lime, the amount of impurities and the particle size distribution. Only three different limes were fully tested and characterised. Six additional Chinese limes were tested for impurities, the lime content and reactivity. The three limes that were fully tested were they from Poland, Sweden, a lime from Partek Nordsjökalk, and one from China The tests for available lime showed on a distinct difference between the limes. Table 1 Content available lime
Type Type of lime lime Avai Availab lable le cont conten entt (%) (%) Swedish Chinese Polish
91 78 66
To test the porosity a whole varity of variables were tested since it was not possible to measure the porosity directly. The measurement made were the specific surface, the pore size distribution, the pore volumes and the density. Table 2 Specific surface (BET)
Type Type of lim limee BETBET-su surf rfac acee (m (m2 /g) Swedish Chinese Polish
1.15 2.67 1.10
1
Table 3 Average pore size (BJH)
The particle size distribution only show on one thing that there was close to no difference between the Swedish and Polish lime considering particle size. The particle size distribution also showed that the Chinese lime consists of particles of a larger size than the two other limes. This might partially explain why the Chinese lime has a higher density than the Swedish lime. Examination of the surfaces performed with XPS gave no clues whether there were any impurities on the surfaces. XPS only showed the existents of only three components on the surfaces carbon, calcium and oxygen.
Type of lime Pore diameter (Å) Swedish Chinese Polish
144 200 182
Table 4 Pore volumes (BJH)
Type of lime Pore volume (cm3 /g) Swedish Chinese Polish
0.0035 0.0118 0.0044
Table 5 Density
Type of lime Density (kg/m3) Swedish Chinese Polish
1210 1250 1300
Table 7 Surface concentration (atomic fraction)
Since quicklime belongs to the bulk chemicals it is difficult to measure the density. The density is therefor given as the figure for the weight per volume when the lime no longer could be compressed through shaking. The figures listed gave no real answers to how reactive each lime would be compared to the others. The density implies that the Swedish lime should be much more reactive than both the polish as well as the Chinese lime. The measured specific surface and the measured pore volumes indicate that the Chinese lime should be much more reactive than both the Swedish and the Polish lime, which should show almost the same reactivity. The average pore size for the different lime gave no new light to which lime that is the most reactive. Table 6 Particle size distribution
<2.83 mm <1 mm <500 µm <250 µm <125 µm <63 µm
Swedish
Chinese
Polish
1 0.81 0.72 0.61 0.47 0.42
1 0.61 0.43 0.34 0.24 0.21
1 0.82 0.71 0.61 0.50 0.40
Swedish Chinese Sample A Sample B Sample C Sample D Sample E Sample F
Calcium Carbon
Oxygen
0.152 0.177 0.157 0.155 0.141 0.118 0.137 0.124
0.671 0.677 0.667 0.634 0.677 0.505 0.575 0.641
0.177 0.146 0.175 0.211 0.183 0.377 0.288 0.235
Samples A through F are the six Chinese lime types. The surface concentrations gave no real new light to how reactive each lime is supposed to be. Apart from the fact that limes with more calcium on the surface perhaps will be more reactive, it is still hard to say anything about the lime reactivity. Since the surface is only a little part of the lime and it is probably not representative for the rest of the lime. What was noticed during the characterisation was that there was very little difference between the Swedish and the Polish limes. Wet slaking
The wet slaking was carried out in an insulated beaker containing 400 ml water to which 100 g lime was added. The
2
temperature rise was measured and this was correlated to the conversion. The slaking procedure was carried out according to the ASTM-standard C110. Besides that, the initial temperature dependence was tested.
The Swedish lime at different starttemperatures 50
Another way of raising reactivity is by adding calcium chloride. Table 8 Slaking with addition of CaCl 2
additive
none
CaCl2
Type of lime Swedish Chinese Polish
Time (min) 6.33 4.67 17
Time (min) 2.67 2 7.67
25 degreas
C ° (40 f f i30 d p20 m e T10
30 degreas 35 degreas 40 degreas 50 degreas
0 0
5 Time (min)
10
The figures in the table above are the slaking times you get if you add 10 grams of calciumchloride per litres of water. The figures clearly show that the slaking times decreases fairly when you add calciumchloride.
The Polish lime at different starttemperatures
30 C25 ° ( f 20 f i d 15 p m10 e T 5 0
Polish lime with different ammonts of calciumchloride added 25 degreas 30 degreas 35 degreas 40 degreas 50 degreas
0
10
20
) 40 C ° ( 30 f f i d 20 p m10 e T 0
0 g/l 4 g/l
10 g/l
0
30
10
Time (min)
20
20 g/l
30
Time (min)
The Chineese lime at different starttemperatures 40
Figure 2 The Polish lime wet slaked with calcium chloride 25 degreas
C ° 30 ( f f i d 20 p m e 10 T
30 degreas 35 degreas 40 degreas 50 degreas
0 0
5 Time (min)
10
What’s interesting in the graph above is that there is a minimum slaking time when calcium chloride is added and that the less reactive the lime is the more effect has increasing the dose of calcium chloride.
Figure 1 Different limes at different starttemperatures Limes slaked differently
In the figure above we can see big differences in slaking times for the Polish lime to the others and we can also notice that the Chinese lime is a little more reactive than the Swedish lime. We can also notice the slaking time seems to converge with raised start temperature for the slaking. Parts of the fact that the Polish curve increases its reactivity depends on the fact that the time difference between total slaking time and the slaking time according to the ASTM-standard levels out with raised start temperature.
) C ° ( f f i d p m e T
Polish lime
60
Polish lime plus 20 g CaCl2x2H2O Swedish lime
40 20
Polish lime 50 °C
0 0
20
40
Time (min)
Figure 3 The Polish lime compared to the Swedish
3
What you can see in the last figure is that when both raising the start temperature and adding calcium chloride to the slaking water the difference in slaking times between the Swedish lime and the Polish lime decreases dramatically. As with many positive effects there is also a downside. Adding calcium chloride lowers the BET-surface of the slaked product, see Table 9.
were fully comparable. Besides one of the samples all samples had a well definable slaking time according to the ASTMstandard. However one of the samples lacked this cause there were no temperature measurement made in the end besides the end point. Plotting the slaking times against the ratio between calcium and carbon atomic surface concentrations gave an almost perfect linear function apart from the outlier discussed earlier.
Table 9 BET-surfaces
Additive
none
CaCl2
Sort of lime
Surface (m2 /g) 26.5 23.4 18.2
Surface (m2 /g)
Swedish Chinese Polish
The slaking time versus the relationship between calcium and carbon on the surface 25 e m i ) 20 t n 15 g i n m i ( 10 k a l 5 S
9.7 7.9
0
The surfaces in the table above are those for the products slaked at a starting temperature of 25 °C. The third way of raising the limes reactivity is grinding. The Polish lime was ground, the fraction above 125 µm was ground to under 125 µm, when you compare this reactivity to that one for the original lime you notice that reactivity has increased dramatically, as you can see in the figure below
0
0.2
0.4
0.6
0.8
1
Ca/C (atomic concentrations)
Figure 4 The slaking time plotte d against surface concentrations of calcium and carbon
The slaking seems to be divided into two parts one initial part that probably involves bursting of the particles and a second phase that seems to be a part where the reaction goes on, on the surface of the cores of the particles. The data from the wet slaking were fitted to the shrinking core model for a surface reaction.
Polish lime ground ) 40 C ° ( 30 f f i d 20 p m10 e T 0
Polish lime
Polish lime ground
0
10
20
30
Time (min) Figure 3 The Polish lime ground
Shrinking core reaction control for the Swedish lime 0,8
25 degreas
3 / 0,6 1 ^ ) x - 0,4 1 ( 1 0,2
30 degreas 35 degreas 40 degreas
0
50 degreas 0
100
200
300
Time (s)
The six additional samples from China were also slaked. These samples were slaked with another equipment totally according to the ASTM-standard. Since these latter samples were slaked with the same equipment and at the same time they
Figure 5 Shrinking core reaction control for the Swedish lime
The model was valid for conversions between 0.4 and 0.9. The figures were adjusted for the different temperatures and
4
the temperature dependence was calculated according to the arrenhius equation and gave the activation energies for the slaking processes, 38 kJ/mol for the Swedish lime, 56 kJ/mol for the Polish and 60 kJ/mol for the Chinese lime. Dry slaking
The limes from Sweden and Poland were also slaked in a two-step dry slaker. The slaker could be modelled as a tank followed by a tube. What was measured was the specific-surface and slaking degrees of the slaked products. The slaking degrees were measured with a TGA, where the slaking degrees could be calculated from the weight loss when the lime was heated above 600 °C, a temperature at which lime decomposes into quicklime and water. It was showed that the slaking degrees for the Polish lime were close to those for the Swedish lime. The slaking degrees are
) g 40 / 2 m30 ( e c a 20 f r u s - 10 T E 0 B
Swedish lime 5.2 kg/h 6.7 kg/h 7.8 kg/h 3.7 kg/h
0
0,2 0,4 0,6 Relationship lime/water (kg/kg)
0,8
correlated to the content available lime. Polish lime ) 30 g / 2 25 m ( 20 e c a 15 f r u 10 s T 5 E B 0
In the figure above you can clearly see how the specific surface of the slaked product depends on the water content in reactor. This result was not dependent of the lime quality. Slaking with addition of Triethanolamin was tested for the Swdeish lime. The major difference was that BET-surface doubled and that the amount of small pores in the product increased. Results and disscussion
It was showed that the only really good way of measure and characterise lime is by measuring the relationship between calcium and carbon on the surfaces. The correlation between the reactivity and the different surface concentrations may depend on some sort of mass transport problem associated with the amount of carbon on the surfaces or it may somehow be connected to the fact that there is much carbonates in the lime. It has further been shown that the wet slaking times for different limes can be decreased by raising the temperature or by adding calcium chloride or by doing both. This temperature dependence behaviour may explain why there is only small differences between slaking degrees in the dry slaking. It’s also suggested that the differences in specific surfaces in the dry slaking experiments arise from the fact that the temperature is lowered when extra water is added and that this might have an positive influence on the specific surface. Literature cited
5.85 kg/h 4.12 kg/h 5.2 kg/h
0
0,2
0,4
0,6
0,8
Relationship lime/water (kg/kg)
Figure 6 The specific surface plotted against the water dosation
ASTM C 110-87 (1987) Standard T est M ethods för Physical Testing of Quick lime, H ydrated L ime and L imestone Babatschev, G.,Kassabova, M. (1969) E influss von T emperatur und E lek trolyten auf die H ydration von ungelöschtem Kalk , Zement-Kalk-Gips (22) 312-316 Becker, H., Zander, Von H. (1976) U ber die N eutralisationsgeschwindigkeit von nass oder trock en
5
gelöschten Kalk hydraten in A bhängigk eit von deren H erstellungsbedingungen, Zement-Kalk-Gips (29) 381-387
Becker, H., Zander. Von H. (1977) U ber die L ösungsgeschwindigk eit von Kalk hydraten, ZementKalk-Gips (30) 287-292 Boynton, R. S. (1966) Chemistry and technology of lime and limestone, 1:a upplagan, John Wiley & Sons Campbell, I. M. (1988) Catalysis at surfaces, 1:a upplagan, Chapman and Hall
Ritchie, I. M., Bing-An, X. (1990) T he kinetics of lime slak ing, Hydrometallurgy (23) 377-396 Schlegel, E., Werner, W., H artmann, H-J. (1976) Z ur H ydrationsgeschwinigk eit von CaO , Silikattechnik (27) 377-378 Zeilnhofer, J., Ploetz, C. (1998) V ollautomatische Kalk -Trock enlöschanlage, Zement-Kalk-Gips International (51) 494-499
R eceived for review february 5, 2001
Devismes, von R., Foster, P., Perraud, R. (1990) E ntschwefelung von schwefeldioxidreichen rauchgasen mit branntkalk und kalk hydrat , Zement-KalkGips (43) 38-42 Frank, G., Achenbach, G. (1987) E influss von Chlorid-Ionen auf das L öschen von Kalk , ZementKalk-Gips International (40) 479-482 Frank, G. (1977) E influss der L öschbedingungen auf die Qualität des gebildeten Calciumhydroxids beim N asslöschen von Kalk , Zement-Kalk-Gips (30) 34-39 Giles, D. E., Ritchie, I. M., Bing-An, X. (1993) T he k inetics of dissolution of slak ed lime , Hydrometallurgy (32) 119-128 Schmitz, F., Hennecke, H. P., Bestek, H., Roeder, A. (1984) Trock engelöschtes Kalk hydrat mit grosser Oberfläche-E in wirk sames reagenz zur bindung saurer abgasbestandteile T eil1: Herstellung im labormaβstab und ausblick über die verwendung bei der rauchgasreinigung, Zement-Kalk-Gips International (37) 530-533 Hennecke, H. P., Köning, W., Roeder, A., Schmitz, F., Stumpf, T. (1986) T rock engelöschtes Kalk hydrat mit grosser Oberfläche-E in wirk sames reagenz zur bindung saurer abgasbestandteile Teil2: A ufbau und betrieb der k leinproduktionsanlage; betriebsergebnisse aus versuchen z ur trock ensorption von schadstoffen aus unterschiedlichen abgasen , Zement-Kalk-Gips International (39) 251-258 Ingesson, E (2000) H ydrox id OH - sock ermetoden Johnson, W. A., Mehl, R. F. (1939) Trans AIME (135) 416 Levenspiel, O. (1972) Chemical reaction engineering, 2nd upplagan, Wiley international editions Ohnemüller, W., Hupe, B. (1969) D ie H ydration des Branntk alk es in der Kalk sandstein-Rohmasse und ihre Bedetung für die Steinfestigk eit vor und nach dem Dampfhärten, Zement-Kalk-Gips (22) 116121
6
