Available online at www.sciencedirect.

com

ScienceDirect
Procedia Engineering 172 (2017) 157 – 164

Modern Building Materials, Structures and Techniques, MBMST 2016

Methodology of sample selection for study of limestone

decarbonation
Jaroslav Bureša, Dušan Dolákb*, Karel Dvořákb, Dalibor Všijanskýc
a
Lime Business Consulting s.r.o, Pekárenská 330/12, Brno 602 00, Czech Republic
b
Brno University of Technology, Faculty of Civil Engineering, Veveri 331/95, Brno 602 00, Czech Republic
c
Masaryk University, Faculty of Science, Kotlářská 267/2, Brno 611 37, Czech Republic

Abstract

Depending on nature of calcination process, we expect to get lime with specific features, for example fast or slow reactivity,
certain grain size or lime that provides hydrated lime with high specific surface area. However, reality is often completely
different. Because of that, there are dozens of works in past 70 years that tried to find correlation between properties of
limestone, calcination parameters and final properties of hydrated lime. None of these papers has managed to find such a relation
in general sense. Our work on this subject starts with proper sample selection. In order to contain the widest range of samples, we
selected only limestones with high purity. All the samples have content of CaCO 3 over 97.8 %. Sample file contains specimens
with various geological age (form 60 mil. to 500 mil. years old), various physico-mechanical properties and different petrology.
These samples are complemented with some limestones with unexpected behavior during calcination process. Rationale behind
selection of each sample is included. We ended up with 26 samples of limestones. These samples were in first phase analyzed
using chemical analysis, mineralogical analysis of insoluble residues and many others to identify differences in behavior. The
results showed differences in the porosity and the associated volumetric mass density. Mineralogy of insoluble residue was in all
cases very similar, so it is not cause of difference in behavior. This means that there are many other factors left for assessment of
effect on behavior of lime, hydrated lime and also limestones during burning process.

©
© 2017
2016The TheAuthors. Published
Authors. by Elsevier
Published Ltd. This
by Elsevier Ltd.is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of MBMST 2016.
Peer-review under responsibility of the organizing committee of MBMST 2016
Keywords: limestone; lime; decarbonation.

* Corresponding author.
E-mail address: dolak.d@fce.vutbr.cz

1877-7058 © 2017 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the organizing committee of MBMST 2016
doi:10.1016/j.proeng.2017.02.044
158 Jaroslav Bureš et al. / Procedia Engineering 172 (2017) 157 – 164

1. Introduction

Lime is one of the oldest building materials known, dating back over 3000 years ago. Historically, there were not
very high demands on its properties. Growth of industrial production, particularly steel production and later growth
of industrial chemistry and modern building material began to emerge higher requirements on specific properties of
lime, particularly with regards to its reactivity. Similarly the requirements for strength of lump lime became higher
to reduce decrepitating and thereby reduce dustiness during manipulation. However it appeared that not all
limestones can produce lime with required properties. Some limestones, even those with high purity can quickly
became hard burnt, others are unable of hard burn despite of excess of burning energy. Some limestones fall apart
during burning, other are willing to sinter. Similar variance is in properties of burnt lime. This issue is significant
and it bothers both manufacturers and users. Because of it, there were a lot of studies and publications within past 50
years in field of limestone, their decarbonation and properties.
Wuhrer (1953, 1956) studied problems of decomposition of limestones during burning process [1,2]. He was first
to describe penetration lines of dissociation in a thin layer from surface to the center of the sample. He also studied
size of pores and their total volume on various limestones, but failed to discover clear dependence to process of
dissociation. Penetration rate of dissociation lines was described by Hedin (1956) [3]. His study shows that
dissociation rate is controlled by escape speed of CO 2 to surface. This effect was described in more detail by Cremer
and Nitsch in 1959 [4]. Following these works, Kainer et al. (1982, 1986) attempted to determine coefficient rate of
dissociation reaction base on diffusion coefficients relative to porosity of limestone [5,6].
Great attention was also given to disintegration of limestones during firing process, because it is significant
technological problem when using shaft furnace. This topic was summarized by Schiele and Behrens (1972), who
found correlation between crystal size and disintegration of limestone [7]. Extreme cases of disintegration are seen
when firing marbles, which are perfectly crystalline.
There was effort to describe sintering process in the end of 19th century, with no result. Clark (1940), Wuhrer
(1953) and Pohl (1963) were first to start systematic study of growth of CaO crystals with increasing temperature
and correlation with density and reactivity of lime [8,9,1,10]. Sintering process can take place even when there is
still unfinished decarbonation inside of sample, as was demonstrated by Pohl (1956) [11]. He also confirmed that
when sintering occurs, density of sample increases with temperature. Temperature of core of samples, which were
not decomposed, was in all cases lower than the decomposed limestone. Lehmann and Schmidt (1961) recount
shrinkage of samples depending on the sintering process [12]. They also distinguish decarbonated lime and sintered
lime, but it is very difficult to separate one from another, because both processes take place almost simultaneously.
Hogewonning and Wolter (2004) summarize in their work all of the existing knowledge of most important
historical studies in this field [13]. They also made complex study of limestones from all 23 existing lime plants in
Germany themselves. They used latest analysis technologies and focused on petrographic and physico-chemical
properties. They exposed their samples to various temperatures and monitored various parameters of limes and
limestones with effort to create mathematical models. These are the most significant findings:

x Sintering process was observed in all samples from 1050°C with one hour of endurance and it is dependent on
specific surface of limestone
x Limestones with larger specific surface area have, because of larger reaction area and therefore lower diffusion
coefficient, higher potential for hard burning.
x Limestones can be, based on characteristics of produced limes, divided into three group.
1. Limestones with higher percentage of impurities and thus lower content of active CaO. This group exhibits
certain stopping of sintering process. Impurities inhibit diffusion process of solids at the interface between
individual crystallites of CaO.
2. Limestones with moderate amount of impurities (especially K 2O and MgO) create mainly limes with small
grain size and with crystal structure of CaO penetrated with impurities.
3. Third group consists from highly pure limestones. In these, diffusion of solid particles is not obstructed and
because of that is sintering process very fast. These limestones have higher prerequisite to create hard burnt
lime.
Despite all the work performed, we couldn’t find sufficient answer why some limestones behave differently than
 Jaroslav Bureš et al. / Procedia Engineering 172 (2017) 157 – 164 159

others. Also, some of the conclusions of work of Hogewonning and Wolter (2004) cannot be generalized [13]. We
want to mainly test their statements on limestones with high saturation factor (higher than 104). One of them is that
reactivity of lime after sintering process is function of density. There known limestones that did not have this
connection. Other statement is that high purity limestones sinter faster and that leads to increased reactivity. There
are also known exceptions for this statements so it cannot be generalized. We decided to try to confirm or refute
mentioned statements in relation to industrial processes and find explanation for some of anomalous phenomena in
the behaviour of limestones during burning process. This paper covers sample selection, initial analysis of chemical
and petrographic composition as well as measurements of porosity and volumetric mass density.

2. Material and methods

We wanted to avoid some parameters during sample selection that may have been, potentially, influential in
previous works:

x Excessive heterogeneity in the amounts of impurities may cause complication in overall assessment. We want to
work exclusively with limestones of high content of CaCO3 > 97 % and with content of carbonates >99 %.
x Samples were collected to also characterize geological age from Paleogene to Precambrian.
x Samples with anomalous behavior are also included. For example samples which are disintegrated even while
soft burning and with low reactivity with t60 > 10 minutes, or samples which cannot be hard burnt even with
extensive temperature.

This way we defined 26 samples of limestones, which are located in Europe, North and South America, Asia.
About 3/4ths of these samples are currently processed into lime and therefore there is experience with industrial
processing and operational characteristics. Each sample was collected by trained and specialized geologist on one
place and then normalized to same particle size for subsequent research. Reasoning behind inclusion of each sample
is included in discussion section.
Chemical analysis was performed by collection of methods, including thermogravimetric method, X-ray
fluorescence and atomic absorption spectroscopy. Petrographic assessment was made on 100 um thick section using
optical microscope. After that, 1.5 kg of each sample was dissolved in 20 % acetic acid at room temperature. This
process ended by complete dissolution of all carbonates. Insoluble residue was filtered, dried and ground prior to
further analysis. Mineralogy of insoluble residues was detected by multifunctional diffractometer Panaytical
Empyrean XRD with anode Cu,Kα radiation. The porosity of samples was determined by high pressure porosimeter
Thermo Finnigan Pascal 240 with a maximal pressure of 200 MPa and pore resolution from 0.0037 to 7.5 μm.
Volumetric mass density was determined by hydrostatic measurements in isopropyl alcohol. Isopropyl alcohol was
chosen instead of water, because in further research, we want to use this method for lime.

3. Results

Categorization of samples by their geological age and structure, their Dunham classification and part of chemical
analysis (content of CaCO3 obtained by loss on ignition method) can be found in Table 1.

Table 1. Classification of limestones - age, chemistry and Dunham classification.

Sample Geological age Category Dunham Classification CaCO3
1 Upper Cretaceous Soft, porous Bioclastic wackestone 98.22 %
2 Upper Jurassic Soft, porous Oolitic grainstone 99.04 %
3 Upper Jurassic Soft, porous Bioclastic lime-mudstone to wackestone 99.27 %
4 Upper Jurassic Soft, porous Oolitic packstone to grainstone 98.38 %
5 Upper Jurassic Soft, porous Oolitic-peloidal grainstone 98.20 %
6 Paleocene Soft, porous Bioclastic packstone-wackestone 98.31 %
160 Jaroslav Bureš et al. / Procedia Engineering 172 (2017) 157 – 164

7 Upper Jurassic Soft, porous Lime mudstone 98.70 %

8 Middle Triassic Medium hard Lumpy packstone-grainstone and bacterial bindstone 98.64 %
9 Lower Cretaceous Medium hard Peloidal-intraclastic packstone-grainstone 98.25 %
10 Mississippian Medium hard Ooidal-bioclastic grainstone 98.72 %
11 Middle Jurassic Medium hard Peloidal grainstone-packstone 97.88 %
12 Middle/Upper Devonian Hard, compact Bioclastic grainstone-cementsone 99.00 %
13 Mississippian Hard, compact Ooidal grainstone 99.25 %
14 Mississippian Hard, compact Bioclastic grainstone-packstone 99.02 %
15 Middle Devonian Hard, compact Bioclastic grainstone-packstone 98.95 %
16 Lower Devonian Hard, compact Bioclastic grainstone 98.59 %
17 Lower Devonian Hard, compact Bioclastic grainstone 97.41 %
18 Mississippian Hard, compact Oolitic grainstone 98.77 %
19 Mississippian Hard, compact Bioclastic packstone-wackestone 98.81 %
20 Lower Devonian Hard, compact Peloidal-bioclastic grainstone to rudstone 98.54 %
21 Mississippian Hard, compact Intraclastic packstone 98.86 %
22 Cryogenian Hard, recrystallized Peloidal cementstone 98.47 %
23 Cryogenian Hard, recrystallized Sparstone to microsparstone 98.25 %
24 Carboniferous/Permian Hard, recrystallized Sparstone 98.81 %
25 Ordovician Hard, recrystallized Sparstone 98.56 %
26 U. Triassic/L. Jurassic Calcite marble Sparstone 97.97 %

Samples can be divided to 4 groups. Samples 1 to 7 are soft and porous. Samples 8 to 11 are medium hard.
Samples 12 to 21 are compact and hard. Last 5 samples (22 to 26) are hard and recrystallized, while sample 26 is
clear marble. Dunham classification ranges from wackestone through grainstone to sparstone, so the range is really
wide. Lowes purity of limestone is 97.41 %, so all samples are incredibly really high percentage limestones.
Mineralogy of insoluble residues of all samples can be find in Table 2.

Table 2. Classification of impurities in limestones - mineralogical analysis.

Sample Mineralogical analysis of insoluble residues
1 Quartz,Fluorapatite, Illite, Montmorillonite
2 Quartz, Illite, Goethite
3 Quartz, Muscovite, Kaolinite, Gypsum
4 Quartz, Illite, Sepiolite,Orthoclase
5 Quartz, Illite, Orthoclase
6 Quartz, Illite, Montmorillonite, Fluorapatite, Orthoclase
7 Quartz, Illite, Kaolinite, Goethite
8 Quartz, Illite, Kaolinite, Muscovite
9 Quartz, Illite, Kaolinite, Montmorillonite
10 Quartz, Muscovite
11 Quartz, Muscovite, Kaolinite, Gypsum, Paragonite, Sepiolite, Goethite, Montmorillonite, Vermiculite
12 Quartz, Muscovite, Dolomite, Illite
13 Quartz, Muscovite, Gypsum, Illite
14 Quartz, Muscovite, Kaolinite, Montmorillonite,Albite
15 Quartz, Muscovite, Illite, Hematite, Orthoclase
 Jaroslav Bureš et al. / Procedia Engineering 172 (2017) 157 – 164 161

16 Quartz, Muscovite, Kaolinite, Gypsum, Pyrite, Fluorapatite, Goethite, Vermiculite, Anorthoclase

17 Quartz, Muscovite, Fluorapatite, Kaolinite, Amesite, Orthoclase
18 Quartz, Goethite, Muscovite (2M1, 2M2)
19 Quartz, Goethite, Muscovite, Kaolinite,Gypsum
20 Quartz, Muscovite, Fluorapatite, Gypsum
21 Quartz, Muscovite, Pyrite, Gypsum, Orthoclase
22 Quartz, Muscovite, Kaolinite, Apatite, Chlorite, Biotite
23 Quatz, Muscovite, Montmorillonite, Albite, Fluorapatite, Pyrite, Fluorite
24 Apatite, Muscovite, Montmorillonite – Chlorite, Goethite, Sylvine, Albite
25 Quartz, Muscovite, Illite, Montmorillonite, Kaolin, Dolomite, Apatite, Pyrite, Orthoclas
26 Quartz, Muscovite, Apatite, Albite, Hydrotalcite

Table with insoluble residues shows that all samples have very similar content of impurities from mineralogical
view. Most common minerals are quartz, feldspar, mica and clay minerals.

Fig. 1. Total porosity of limestone samples.

Results show that samples 1 to 8 and also 11 have relatively large volume of pores. All other samples are
compact with very small amount of pores.
162 Jaroslav Bureš et al. / Procedia Engineering 172 (2017) 157 – 164

Fig. 2. Volumetric mass density of all samples.

Volumetric mass measured by hydrostatic method show that majority of samples have density over 2200 kg·m-3.

4. Discussion

Based on previous works in similar field, for our study of behavior of limestones during burning process in
temperature range of 900°C to 1360°C we have chosen limestone samples which are (for the most part) used in lime
production industry, so there is experience in their operational usage and also are known properties of limes made
from them. Tests of thermal load will be part of further research. Unlike previous works, all chosen samples are high
percentage limestones (CaCO3 content bigger than 97%) and have different geological ages, from Precambrian
(500-600 mil. Years old) to Paleogene (60 mil. Years old). Samples with other than predicted behavior are also
included.
Samples can be divided to 4 groups, based on porosity and hardness. Samples from 1 to 7 are soft and porous.
Samples from 8 to 11 are medium hard. Compact and hard samples are from 12 to 21. Last group are samples from
22 to 26, which are hard and recrystallized. Sample 26 is marble.
Diversity of sample file is significant, but all samples are similar in their purity. First sample is chalk from France
and represents younger, porous limestones form Upper Cretaceous era. During calcination, lump lime is hard and
compact enough, so it does not crumple. Samples 2 and 3 are from Poland and from same deposit, just from
different horizons. Main difference between them is in their crystal structure and also in properties of lime hydrate.
Both are from Upper Jurassic era and are wet and porous. Samples 4 and 5 are from one deposit in France. They are
also from Jurassic era, but are different from 2 and 3, mainly in their hardness. Difference between themselves is in
properties of lime hydrate. Sample 6 is from Paleocene era and is mined in Denmark. It is also chalk as sample 1,
but it is not as hard during calcination, so it decrepitates. Sample 7 is similar to samples 4 and 5, it is from same era
and county, but properties of hydrated lime are completely different. Sample 8 is Middle Triassic limestone from
Poland. This sample is prime example of readily calcinating limestone. It is possible to create hard burnt lime event
in furnace for soft burn. Sample 9 is limestone from Lower Cretaceous period and is mined in France. It is used for
production of soft burnt lime and during calcination doesn’t decrepitate. Sample 10 is from Mississippian era and it
is mined in USA. It is also used for production of soft burnt lime, but it does decrepitate during calcination, but not
 Jaroslav Bureš et al. / Procedia Engineering 172 (2017) 157 – 164 163

to much extend (like samples 6 or 16). Sample 11 is from Middle Jurassic era, originates in Iberian area and has
standard properties. Sample 12 if from Germany and is form Middle/Upper Devon. It is easily calcinated, does not
crumple. Sample 13 is from Belgium, it is from Mississippian era and is similar in its behavior during calcination to
12. It does not decrepitate and is used for production of soft burnt lime. Sample 14 interesting limestone form
Mississippian era mined in UK. Its behavior is not standard, in furnace for soft burn creates hard burnt lime. Also,
hydrated lime has really specific properties. Sample 15 is Middle Devonian limestone from Denmark. It is easily
calcinated. Samples 16 and 17 are from same deposit in Czech Republic, but from different horizons. They are from
Lower Devon and have different structure and properties. Especially sample 16 is unique in its behavior. During
calcination decrepitate extensively, it is not possible to create hard burnt lime under normal circumstances and lime
hydrate has lowest value of sedimentation volume. Sample 19 is from Mississippian era and is mined in France. It is
used for production of soft burnt lime and lime hydrate has standard properties. Samples 22 and 23 are from
Cryogenian era, both from Brazil, but from different deposits. They represent oldest limestones. Sample 24 is from
Permian era and is mined in Malaysia. It is used for production of soft burnt lime. Similarly to sample 14, its lime
hydrate has some really specific properties. Sample 25 is from Norway and was created in Ordovician era. It is other
example of really old limestone. Properties of lime hydrate are standard, but lump lime decrepitates during
calcination.
Samples 18, 20, 21 and 26 are not used for production of lump lime or lime hydrate, so we do not have
description of their behavior during calcination. Nevertheless, each of these samples have their place in sample file
because of their properties. Sample 18 is from Belgium, it is from Mississippian era and was chosen because its
interesting crystal structure. Sample 20 was chosen because of combination of geological age and locality. It’s the
only Devonian limestone from Russia. Sample 21is from Mississippian era from Belgium and it was chosen
because its incredible hardness. Sample 26 is from Italy, it was created during Upper Trias/Lower Jurassic and it’s
the only true marble in sample file.
Petrographic description of samples is shrinked for purpose of this paper to contain only Dunham classification.
As seen in table 1, variety of samples is immense. There are samples of wackestone, grainstone, sparstone,
packstone and others. Porosity of samples ranges from 0.5% to almost 44%. Most of the samples are compact with
porosity lower than 5%. Volumetric density of samples ranges from 1300 kg·m-3 to 2700 kg·m-3. X-ray diffraction
analysis of insoluble residues shows content of quartz, calcite, mica (muscovite), feldspar (albite) and clay minerals
as are kaolinite, illite and montmorillonite in all samples. These minerals are in internal structure of limestones and
were inserted during actual formation by water in form of fine particles. Furthermore some samples had content of
specific minerals, as is, for example fluorapatite (1,6,16,17,20,22,23,25,26), goethite (2,11,16,18,19,24), orthoclase
(5,6,15,16,17,21,25), pyrite (16,21,23,25), gypsum (3,11,13,16,19,20,21), dolomite (12, 25), fluorite (23), hematite
(15) and the hydrotalcite (26). Calcite was detected in some samples as residue after dissolution. Mineralogical
composition of insoluble residues of all sample was similar, regardless of age or type of limestone. No heavy metals
were detected, so they are non-existent or below detection limit. Also, calcium acetate was not detected in any of
samples, thus filtering of insoluble residues was ideal.
Results of mineralogy of insoluble residues, as well as porosity don’t provide any answers to unusual behavior of
limestones. Parts of behavior of some limestones can be justified by their crystal structure or porosity. For example,
crumbling of sample 25 can be assigned to its crystal structure. There is prediction that sample 26 will be also
decrepitate under thermal load, because its full crystal structure. Because of porosity of samples from 1 to 7 we can
assume that decalcination speed will be higher than other samples.
Taking to account previous work of Hogewonning and Wolter (2004) [13], we wanted our further methodology
to by closer to their process in order to compare the results in the future and even jointly assess them. Because of
this, our next work is divided into three parts:

x Study of decomposition of limestone by using differential scanning calorimetry (henceforth as DSC), differential
thermal analysis (henceforth as DTA) and porosimetry
x Study of sintering process of CaO in wide range of temperatures and monitoring of changes in microstructure
using scanning electron microscope (henceforth as SEM), BET (using Braunauer-Emmett-Teller theory) and BJH
analysis (using Barrett-Joyner-Halenda theory).
164 Jaroslav Bureš et al. / Procedia Engineering 172 (2017) 157 – 164

x Study of hydration process and properties of hydrated lime using BET and laser granulometry.

The effort of future work is to interpret mathematical models from work of Hogewonning and Wolter (2004) on
our results [13]. Wide range of geological age, porosity, operational characteristics and other properties of
limestones and products made from them, while maintaining high purity of samples, create solid base for further
research.

4. Conclusion

By specific sample selection we created different conditions for study of behavior of limestones during burning
than previous works. We based on operational experience in production of lime in different location in world and
combined limestones with predictable behavior with limestones with other than predicted properties of final lime. In
addition, all samples are similar in their high purity, to minimalize influence of impurities on the evaluation of
results. Also, samples were selected to include wide range of geological periods, in which limestones are found.
Because of this, we created unique set of samples that will be used for a very comprehensive study. Initial analysis
of insoluble residues do not show any sign of major differences between samples, so we can say that there must be
another parameter or combination of parameters which are responsible for unusual behavior of limestones.

Acknowledgements

This work was financially supported by project number: 15-08755S “Study of effects of samples preparation on
inorganic binders final properties”, project number FAST-J-16-3734 “Study of influence of limestone properties on
final properties of lime” and project No. LO1408 "AdMaS UP - Advanced Materials, Structures and Technologies",
supported by Ministry of Education, Youth and Sports under the „National Sustainability Programme I".

References

[1] J. Wuhrer, J. Zement-Kalk-Gips International 6 (1953) 354-368.

[2] J.Wuhrer, G. Rademacher, W. Lahl J. TIZ International 80 (1956) 100-104.
[3] R.Hedin, J. Zement-Kalk-Gips International 9 (1956) 521-524.
[4] E.Cremer, J.Nitsch, TIZ International 83 (1959) 579-586.
[5] H.Kainer, E.Specht, R. Jeschar, Pore diffusion, reaction and thermal conduction coefficients of various linestones and their influence on
decomposition time, Zement Kalk Gips 39(7) (1986) 214-219 (englisch)
[6] R.Jeschar, H. Kainer, E.Specht, Der Einfluß carbonatischer Zuschläge auf das thermische Verhalten von Beton unter Brandbeanspruchung
Schriftenreihe des SFB 148 "Brandverhalten von Bauteilen", TU Braunschweig, 1982.
[7] E. Schiele, L.W.Behrens, Kalk - Herstellung, Eigenschaften, Verwendung, 627 (1972).
[8] F.Birch, H.Clark, The thermal conductivity of rocks and its dependence upon temperature and composition part I, Am J. Sci. 238(8) (1940)
529-558.
[9] F. Birch, H. Clark, The thermal conductivity of rocks and its dependence upon temperature and composition part II, Am J. Sci. 238(9) (1940)
615-635.
[10] G. Pohl, J.TIZ International 87 (1963) 417-424.
[11] G. Pohl, J.Zement-Kalk-Gips International 12 (1959) 566-572.
[12] H.Lehmann, M.Schmidt, J. TIZ International 85 (1961) 73-91.
[13] A. Wolter, S. Hogewoning, Vorhersage und Beeinflussung der Reaktivität von Branntkalk, Forschungabericht für die
Forschungsgemeinschaft Kalk und Mörtel e.V.: 94 S, TU Claustahl. (2004).