Thermal decomposition study of crystalline limestone

Construction and Building

MATERIALS

Construction and Building Materials 19 (2005) 359–365

www.elsevier.com/locate/conbuildmat

Thermal decomposition study of crystalline limestone using P-wave velocity N. Kantiranis a

a,*

, A. Filippidis a, A. Tsirambides a, B. Christaras

b

Department of Mineralogy–Petrology–Economic Geology, School of Geology, Aristotle University of Thessaloniki, Thessaloniki GR- 541 24, Greece b Department of Geology, Laboratory of Engineering Geology and Hydrogeology, School of Geology, Aristotle University of Thessaloniki, GR-541 24, Greece Received 7 February 2003; received in revised form 30 July 2004; accepted 2 August 2004 Available online 25 September 2004

Abstract A high-calcium limestone was calcined in order to study its thermal decomposition using P-wave velocity. The onset of calcination is at approximately 750 C, while its completion according to the size of the cubic specimens is noted between 1000 and 1150 C. We found that P-wave velocities are a very good index for the study and estimation of calcination. P-wave velocity decreases due to a temperature rise from 650 to 1150 C for the cubic specimens of 4, 6 and 8 cm mean edge, while for 1 and 2 cm cubic specimens mixed behavior is observed, with a considerable increase in velocities at calcination temperature higher than 1050 C. 2004 Elsevier Ltd. All rights reserved. Keywords: Limestone; P-wave velocity; Thermal decomposition

1. Introduction The main chemical property of limestone is its thermal decomposition, known as ‘‘calcination’’, during which quicklime (CaO) and carbon dioxide are produced. It is a reaction that is strongly affected by the partial pressure of the gas phase (PCO2). An increase in PCO2 partial pressure leads to a rise in the initial calcination temperature [1,2]. According to Boynton [1] CaCO3 decomposition temperature determined by several researchers at the beginning of the 20th century are generally still accepted as 898 C at 1 atm in a 100% CO2 environment. However, according to some studies, this temperature is set at 902.5 C [3–5]. The thermal decomposition of MgCO3 (magnesite) is performed at much lower temperatures between 402 and 550 C depending on the CO2 partial pressure [1,6,7]. Since the MgCO3/CaCO3 ratio varies with the type of

limestone, the decomposition temperature does not remain constant and therefore must be determined for every type of limestone. Moreover, differences in the crystallinity of the limestone seem to consolidate the discordance of data. The decomposition of MgCO3 of dolomite is performed at higher temperatures than magnesite. A complete decomposition average value of a high-crystallinity dolomite at 1 atm pressure and 100% CO2 environment is set at 800 C [1]. The CaCO3 of dolomite resists at higher temperatures, resulting in two decomposition stages. The aim of this study is to examine the calcination process using P-wave velocity of cubic specimens of several sizes from specific crystalline limestone subjected to calcination experiments under various temperature conditions and retention times. 2. Materials and methods

*

Corresponding author. Tel.: +30 2310 998437; fax: +30 2310 998463. E-mail address: [email protected] (N. Kantiranis). 0950-0618/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.conbuildmat.2004.08.002

Cubic specimens of 1, 2, 4, 6 and 8 cm mean edge were prepared from crystalline limestones collected in

360

N. Kantiranis et al. / Construction and Building Materials 19 (2005) 359–365

the area of Agios Panteleimonas in Florina (NW Greece). More specifically, 72 cubes of 1 cm mean edge (average value 1.14 ± 0.06 cm), 71 cubes of 2 cm mean edge (average value 2.01 ± 0.09 cm), 56 cubes of 4 cm mean edge (average value 4.08 ± 0.13 cm), 56 cubes of 6 cm mean edge (average value 6.19 ± 0.16 cm) and 43 cubes of 8 cm mean edge (average value 8.07 ± 0.27 cm) were prepared. All 298 cubes of approximately 101 kg total weight were calcinated. According to Kantiranis [8] the average size of the maximum dimension of the studied limestonesÕ calcite grains is 0.9 mm, their texture is holocrystalline, and by the XRD estimation of their mineralogical composition, it was found that they are very pure, with an average calcite content of 98%. Furthermore, they contain dolomite as well as traces of micas, quartz, chlorite, feldspars, hematite, amphiboles, epidote, pyrite, titanite, ilmenite and zircon. The organic matter is measured at average value of 0.8%, while the insoluble residue of 1.0%. The experimental calcination was performed at temperatures of 650, 750, 850, 950, 1050 and 1150 C. Retention times (i.e., the time at which limestone fragments remained inside the kiln at the calcination temperature) for every temperature were 60, 120 and 180 min and the preheating rate was 5 C/min in all cases. The calcination experiments were performed in a Naber– Multitherm N11/HR furnace (tmax: 1260 C, ±2 C). All specimens have been weighed before and after their ignition and the difference expressed as a percentage (wt%) is the ignition loss for every calcination condition. The calculation of compression wave (P-wave) velocity passing through a material was conducted according to the French standard AFNOR NF B 10505 [9]. Measurements were made using a PUNDIT velocimeter taking into account every dimension (X, Y, Z) of the specimen. Special uniplanar transducers (see Fig. 1)

are adjusted to the polar edges of the specimen, while for the measurement of P-wave traveling times in this study 54 kHz transducers were used. According to the ASTM C597 [10] for specimens with short path lengths (50 mm or shorter), where the loss of signal is not the governing factor, it is preferable to use vibrational frequencies of 50 kHz or higher to achieve more accurate transit-time measurements and hence greater sensitivity. By dividing the specimensÕ length (traveling distance of P-waves) by the traveling times, we are able to estimate the velocity of P-waves through the cubic specimens. In order to improve the acoustic contact between the specimen and the transducers, yellow plasticine was used as a contact agent, while for the accurate adjustment of the device was equipped with a cylindrical aluminum alloy template (Fig. 1). According to Lemoni [11] plasticine slightly affects the P-wave velocity and the measurements that are received are particularly reliable comparatively with other contact agents, such as aluminium plates, vaseline, grease, soft soap, etc. The calcined specimens were cooled in a dehumidifier and then were measured at ambient temperature. The ultrasonic technique for the estimation of calcination was applied experimentally for the first time by Kantiranis [12] and showed [12,13], that it is a very useful and reliable technique for the study of calcination of a carbonate rock. P-wave velocity is a very good index for the quality of rocks and other materials (e.g., mortars) [14–19].

3. Results and discussion In Table 1 the average P-wave velocity (3 readings/ specimen · 298 specimens = 894 readings), its SD and the average loss of ignition of the samples studied were given. The correlation of the loss of ignition with the

Fig. 1. PUNDIT velocimeter and studied specimens.


361

Table 1 Average P-wave velocity (Vp), SD and average loss of ignition (LOI) of the studied limestone samples Calcination conditions

Cubes 1 cm

Cubes 2 cm

Cubes 4 cm

Temperature (C)

Retention time (min)

Vp (m/s)

SD

LOI (wt%)

Vp (m/s)

SD

LOI (wt%)

Vp (m/s)

SD

LOI (wt%)

Vp (m/s)

SD

LOI (wt%)

Vp (m/s)

SD

LOI (wt%)

650 650 650 750 750 750 850 850 850 950 950 950 1050 1050 1050 1150 1150 1150

60 120 180 60 120 180 60 120 180 60 120 180 60 120 180 60 120 180

2177 2119 2116 1803 1913 1824 1559 1532 1503 1179 1110 936 927 886 903 1042 1228 2045

92 84 110 60 70 76 48 113 77 79 126 139 51 116 111 296 372 1696

0.3 0.2 0.3 0.6 1.2 0.8 6.4 9.5 8.9 31.2 36.5 40.3 42.4 43.2 43.3 43.5 43.7 43.7

2118 2238 2164 1949 1921 1777 1616 1529 1578 1290 1281 1184 834 768 745 807 871 841

74 88 182 101 95 122 81 49 91 114 105 42 77 37 39 69 85 105

0.2 0.3 0.3 0.5 0.7 0.5 3.6 3.3 4.3 18.7 22.5 31.3 42.6 43.1 43.2 43.6 43.6 43.6

2248 2077 2007 1687 1755 1856 1377 1473 1455 1204 1147 1139 962 767 877 762 911 868

208 339 204 270 170 105 159 69 105 106 54 91 80 65 180 84 49 265

0.2 0.3 0.3 0.9 1.3 1.4 4.6 6.3 7.9 17.8 18.4 23.3 40.3 43.4 43.1 43.6 43.6 43.4

1894 2150 2313 1961 1791 1762 1546 1450 1489 1201 1199 1079 950 897 841 741 848 760

311 101 147 149 121 195 100 111 126 83 97 97 61 61 114 56 175 97

0.0 0.0 0.0 0.9 0.8 1.6 4.8 6.9 9.0 17.5 22.5 27.5 36.5 41.6 43.7 43.7 43.7 43.6

2185 2210 2186 1855 1772 1761 1476 1432 1492 1160 1121 1072 1077 1040 945 897 857 753

76 99 122 85 180 120 111 138 119 102 105 59 120 79 106 83 117 100

0.0 0.0 0.0 0.8 1.2 1.1 3.2 5.3 6.7 14.1 17.5 21.0 30.2 36.3 40.2 41.7 43.7 43.9

calcination temperatures is presented in Fig. 2. The sigmoidal or Boltzmann distribution was used for the correlation of the data according to the following formula: y¼

A 1 A2 xx0

1 þ e dx

þ A2 ;

where y represents the loss of ignition (wt%), x is the calcination temperature (C) and A1, A2, x0, dx are the Boltzmann constants. According to Table 1 and Fig. 2 the onset of calcination is at approximately 750 C (average loss of ignition >1%). The completion of calcination, depending on the size of cubic specimens, is noted between 1000 and 1150 C. The cubic specimens of 1 cm mean edge are completely calcined at approximately 1000 C for retention times longer than 120 min, while the cubic specimens of 8 cm mean edge at 1150 C even for retention times of 60 min. For the other specimen sizes, the completion of calcination is observed at intermediate conditions. For cubic specimens of the same edge, calcination at a specific temperature depends strongly on retention time. Extended retention times lead to lower temperatures of complete calcination. Ruckensteiner et al. [20] studied different limestones and found that their calcination begins at 520 C and it is completed at 895 C. Kantiranis [12] and Kantiranis et al. [13] studied cubic specimens of 5 cm mean edge from crystalline limestones of Agios Panteleimonas in Florina and found that their calcination begins at approximately 740 C. On the initial cubic specimens of the studied limestone 801 measurements of P-wave velocity at ambient temperature in all cubic directions were conducted before the calcination. An average value of 5518 m/s with SD 475 m/s was found, with a minimum value of 3643

Cubes 6 cm

Cubes 8 cm

m/s and maximum value of 6427 m/s, while no any specimen size effect was observed. P-wave velocity in limestones ranges between 3500 and 6500 m/s [21], while for a calcite crystal the P-wave velocity is 6490 m/s [22]. The correlation of P-wave velocity (m/s) with the calcination temperature (C) is presented in Fig. 3. For this correlation the third grade polynomial distribution was used. Based on Table 1 and Fig. 3, the average P-wave velocity decreases due to temperature rise for the cubic specimens of 4, 6 and 8 cm mean edge and ranges from 2248 to 753 m/s. Especially for the cubic specimens of 8 cm mean edge, this transition is very smooth. The cubic specimens of 1 and 2 cm mean edge present a more complex behavior and, more specifically, up to 1050 C, P-wave velocity decreases to 745 m/s. The velocity then increases due to an increase of the retention time at calcination temperatures higher than 1050 C. This phenomenon is intense in the case study of the calcination of the 1 cm specimens especially at 1150 C for a retention time of 180 min, during which an average P-wave velocity of 2045 m/s was measured. This increase may be related only to a decrease in porosity and an increase in the dry apparent weight representing the shrinkage of the produced quicklime [8]. According to Rubiera et al. [23] and Wong and Bradit [24], during completion of calcination the textural characteristics of the calcinated specimens are constantly changing due to contraction and shrinkage phenomena. The shrinkage phenomena and the density growth are due to the fact that the CaO molecular crystals formed by the thermal decomposition of CaCO3 are united at the same time into gradually bigger crystals, the ‘‘crystallites’’. In a pure calcite at 900 C these crystals are only approximately 0.1 lm in size, and at

362


Fig. 2. Correlation of the average loss of ignition (wt%) vs calcantion temperature (C) (A1, A2, x0, dx Boltzmann constants, R2 correlation coefficient).

1000 C they are 1 lm in size, due to the symphysis of these crystallites. At 1100 C an even bigger agglomeration of crystallites is formed turning constantly into a more compact mass relative to the temperature rise [25]. Also, an important point is that at an ignition temperature of 650 C and retention time 60 min and despite the absence of calcination phenomena, a considerable decrease in P-wave velocity is noted from an average value of 5518 to 2124 m/s. This behavior is probably related to the structural changes of the lime-

stone during heating. The expansion/contraction of calcite crystals, but also the combustion of the organic material that a limestone may contain leaving microscopic holes that decrease P-wave velocity [1,2,13]. In coarse grain crystalline limestones the instability phenomena of their fragments and fragmentation are due to the different thermal expansion coefficient of the large calcite crystals horizontal (25 · 106 K1) and vertical (5 · 106 K1 ) to the c axis. Limestones intersected by a dense vein system empty or filled with


363

Fig. 3. Correlation of the average P-wave velocity (m/s) vs calcination temperature (C) (R2 correlation coefficient).

sparitic calcite are highly susceptible to fragmentation [20]. P-wave velocity, as a natural characteristic of rocks and different materials, depends on their microand macro-structure, the existence of minor cracks and porosity and the characteristics of their mineralogical components, such as elastic parameters, density and microporosity [22]. Increased velocity with an increase in the dry apparent weight and vice versa is reported by Babus˘ka [26] and Kopf et al. [27]. The effect of minor cracks on the distribution of P-wave velocity in rocks has been studied by Babus˘ka et al. [28] and Jech et al. [29]. The P-wave velocity of a rock decreases with the growth of the average size of grains [22]. Based on the study of P-wave velocity, it is concluded that it depends directly on the calcination temperature and, consequently, the loss of ignition. More specifically (Table 2), a decrease by approximately 61% (5.4 m/s/C) is observed between the average value of P-wave velocity in the initial material and at the milder ignition temper-

ature (650 C). As temperature rises to 750 C and to 850 C, almost the same decrease in P-wave velocity is observed at approximately 15% (3.2 m/s/C) and 18% (3.3 m/s/C), respectively. On the contrary, at temperatures of 950 C, where according to the literature the thermal decomposition of calcite is completed, a more considerable decrease in P-wave velocity is observed of 23% (3.5 m/s/C). At higher temperature of 1050 C, where the calcination of most specimens of the studied limestone is completed, a decrease of 22% (2.6 m/s/C) in P-wave velocity is noted. Finally, at the highest temperature (1150C) P-wave velocities show a mixed behavior. For the cubic samples of 1 and 2 cm mean edge an increase of P-wave velocity is observed, with an average value of 58% and 7%, respectively. This increase probably indicating the contraction and/or shrinkage of the specimens. Almost the same average P-wave velocity is determined for the cubic samples of 4 cm mean edge at temperatures 1050 and 1150 C,

364


Table 2 Percentage variation (%) of the P-wave velocity of the samples studied Calcination temperature (C)

Cubic samples with mean edge 1 cm

2 cm

4 cm

6 cm

8 cm

Initial material ! 650 650 ! 750 750 ! 850 850 ! 950 950 ! 1050 1050 ! 1150

61 14 17 30 15 +58

61 13 16 20 37 +7

62 16 19 19 25 3

62 13 19 22 23 13

60 18 18 24 9 18

Mean value 61 15 18 23 22 –

, Decrease of velocity; +, increase of velocity.

while a decrease is observed, with an average value of 13% and 18% for the cubic samples of 6 and 8 cm mean edge, respectively. Timur [30] has examined the variation of ultrasonic velocity in relation to temperature in specimens of sedimentary rocks maintaining the external and internal pressure constants. He found that P-wave velocity shows a linear decrease with temperature up to 200 C. According to Toksoz et al. [31], an important factor affecting the decrease of P-wave velocity, apart from the dry apparent weight, is the creation of micro-cracks. These micro-cracks are created due to anisotropic thermal expansion of the material inside the cracks, in relation to the matrix, when a rock contains secondary veins.

4. Conclusions The onset of calcination is at approximately 750 C, while its completion according to the size of the cubic specimens is noted at between 1000 and 1150 C. For similar sizes of cubic specimens, complete calcination at a specific temperature depends on the retention time at this temperature. Extended retention times lead to lower temperatures of complete calcination. We found that P-wave velocities are a very good index for the study and estimation of calcination. More specifically, P-wave velocity decreases due to a temperature rise for the cubic specimens of 4, 6 and 8 cm mean edge, while for 1 and 2 cm cubic specimens mixed behavior is observed, with a considerable increase in velocities at calcination temperatures higher than 1050 C. A decrease of approximately 61% (5.4 m/s/C) in the P-wave velocity at 650 C compared to its value in the initial material was observed. As temperature rises to 750 C and to 850 C, almost the same decrease in P-wave velocity is observed of 15% (3.2 m/s/C) and 18% (3.3 m/s/C), respectively. At temperature of 950 C, a more considerable decrease in P-wave velocity of 23% (3.5 m/s/C) is observed. At higher temperature of 1050 C, a decrease of 22% (2.6 m/s/C) in the P-wave velocity is observed. Finally, at the highest tem-

perature (1150 C) P-wave velocities show a mixed behavior. For the cubic samples of 1 and 2 cm mean edge an increase of P-wave velocity is observed, with an average value of 58% and 7%, respectively. This increase probably indicating the contraction and/or shrinkage of the specimens. Almost the same P-wave velocity is determined for the cubic samples of 4 cm mean edge at temperatures 1050 and 1150 C, while a decrease is observed, with an average value of 13% and 18% for the cubic samples of 6 and 8 cm mean edge, respectively. Studying samples during and after the process of calcination at lime kiln by ultrasonic technique in the lime industry, could be a useful tool for the evaluation of the calcination process in industrial scale. Phenomena of incomplete calcination that are caused by soft conditions of calcination (small retention time and/or low temperature), as well as phenomena of hard burning that are owed in extended retention at high temperatures of calcination leading to constriction of the limestone fragments and finally to the reduction of the reactivity of the produced lime, can very easily be determined with the use of this non destructive, practical and rapid method.

References [1] Boynton SR. Chemistry and technology of limestone. 2nd ed. New York: Wiley; 1980. p. 577. [2] Oates JAH. Lime and limestone. Chemistry and technology, production and uses. Weinheim: Wiley–VCH; 1998. p. 455. [3] Turkdogan TE, Olsson GR, Wriedt AH, Darken SL. Calcination of limestone. Trans Soc Min Eng-AIME 1973;254:9–21. [4] Thompson LJ. Predicting lime burning rate via new dynamic calcination theory. Solid state diffusion controls reaction speed in kiln. Pit Quarry 1979;5:80–3. [5] Borgwardt HR. Calcination kinetics and surface area of dispersed limestone particles. AIChE J 1985;31(1):103–11. [6] Schwarzkopf F. Lime burning technology-a manual for lime plant operators. third ed.. Kennedy Van Saun: Svedala Industries; 1994. [7] Gutschick AK. Lime and Limestone. In: Howe-Grant M, editor. Encyclopedia of chemical technology. New York: Wiley; 1995. [8] Kantiranis N. Calcination study of the crystalline limestone from Agios Panteleimonas, Florina, Greece. Ph.D. thesis, School of Geology, Aristotle University of Thessaloniki; 2001. p. 196 [in Greek with extensive English summary]. [9] AFNOR NF B 10505. Produits de carrie`res – Pierres calcaires – Mesure de la vitesse de propagation du son (ondes longitudinales). France Norm; 1973. p. 2. [10] ASTM C 597. Standard test method for pulse velocity through concrete. USA Standard; 1983. p. 3. [11] Lemoni E. Technico-geological classification of soils using a portable ultrasonic meter. Application in the region of KozaniGrevena. M.Sc. thesis, School of Geology, Aristotle University of Thessaloniki; 1998. p. 178 [in Greek]. [12] Kantiranis N. Petrological, geochemical and technological study of the Jurassic carbonate rocks form Agios Panteleimonas, Florina, Macedonia, Greece. M.Sc. thesis, School of Geology,


[13]

[14] [15]

[16]

[17]

[18]

[19]

[20]

[21]

Aristotle University of Thessaloniki; 1998. p. 69 [in Greek with extensive English summary]. Kantiranis N, Tsirambides A, Filippidis A, Christaras B. Technological characteristics of the calcined limestones from Agios Panteleimonas, Macedonia, Greece. Mater Struct 1999;32:546–51. Irfan TW, Dearman WR. Engineering classification and index properties of a weathered granite. Bull IAEG 1978;17:79–90. Auger F. Influence des fluides interstitiels sur la vitesse du son dans les mate´riaux de construction. Mesures expe´rimentales et conse´quences sur les diagnostics dÕalte´rabilite´. Lyon: Intern. Measurm. Testing Civil Eng.; 1988. p. 259–68. Christaras B. Me´thode dÕe´valuation de lÕalte´ration et modification des proprie´te´s mećaniques des granites en Gre`ce du Nord. Bull IAEG 1991;43:21–6. Christaras B. Anisotropy effects on the elastic parameters of rocks; determination using ultrasonic techniques. Bull Geol Soc Greece 1994;4:381–7. Christaras B, Auger F, Mosse E. Determination of the elastic moduli of rocks. Comparison of the ultrasonic velocity and mechanical resonance frequency methods to the direct static one. Mater Struct 1994;27:222–8. Topal T. Ultrasonic testing of artificially weathered Cappadocian tuff. In: Proc. Congr. LCP Õ95. Preserv. Restor. Cultur. Heritage. Montreux;1995. p. 205–12. Ruckensteiner G, Burczeck J, Ludwing U. Factors affecting the stability of the internal structure when burning limestone. Zement-Kalk-Gips Int 1995;6:342–8. Parasnis SD. Principles of applied geophysics. London: Chapman & Hall; 1997. p. 429.

365

[22] Vajdova´ V, Prˇslash ikryl R, Pros Z, Klı´ma K. The effect of rock fabric on P-waves velocity distribution in amphibolites. Phys Earth Planet In 1999;114:39–47. [23] Rubiera F, Faurtes BA, Pis JJ, Artos V, Marban G. Changes in textural properties of limestone and dolomite during calcination. Thermochim Acta 1991;179:125–34. [24] Wong LL, Bradit CR. Lime refractories with limestone and synthetic calcium hydroxide additions. J Am Ceram Soc 1995;78(6):1611–6. [25] Hedin R. Changes occurring in the limestone during heating before calcination. Swed Cem Conc Res Inst Bull 1961;23:34. [26] Babus˘ka V. Elastic anisotropy of igneous and metamorphic rocks. Stud Geph Geod 1968;12:291–303. [27] Kopf M, Mu¨ller HJ, Gottesmann B. Correlation between pyroxene content and VP and VS under high pressure. In: Kapicˇka A, Kropaćˇek V, Pros Z, editors. Physical properties of the mineral system of the earthÕs interior. Prague: Union Czech. Math. Phys.; 1985. [28] Babus˘ka V, Pros Z, Franke W. Effect of fabric and cracks on the elastic anisotropy in granodiorite. Publ Inst Geophys Pol Acad Sci A 1977;6(117):179–86. [29] Jech J, Babus˘ka V, Pros Z. Quantitative correlation of velocity anisotropy with the distribution of microcracks in rocks. In: Kapicˇka A, Kropaćˇek V, Pros Z, editors. Physical properties of the mineral system of the earthÕs interior. Prague: Union Czech. Math. Phys.; 1985. [30] Timur A. Temperature dependence of compressional and shear wave velocities in rocks. Geophysics 1977;42:950–6. [31] Toksoz MN, Cheng CH, Timur A. Velocities of seismic waves in porous rocks. Geophysics 1976;41:621–45.