The role of petrography on the thermal decomposition and...
Transcript of The role of petrography on the thermal decomposition and...
ORIGINAL PAPER
The role of petrography on the thermal decompositionand burnability of limestones used in industrial cement clinker
Nicoletta Marinoni1 & Andrea Bernasconi1 & Giovanna Della Porta1 &
Maurizio Marchi2 & Alessandro Pavese1
Received: 20 October 2014 /Accepted: 30 June 2015# Springer-Verlag Wien 2015
Abstract The present research examines the influence of thepetrographic features on the thermal decomposition andburnability of three limestones, the main raw materials forPortland cement-making. A detailed characterisation of thelimestones has been performed by means of optical microsco-py and X-Ray Powder Diffraction. The carbonate thermaldecomposition was conducted under isothermal conditionsby means of in situ High Temperature X-Ray Powder Diffrac-tion and the heated samples were further investigated by Scan-ning Electron Microscopy. Three kiln feeds were then pre-pared and submitted to burning trials and the temperature ofoccurrence of the main clinker phases was investigated as wellas the content of the uncombined CaO in the heated sampleswas determined by using the Franke method. The results attestthat the microfabric, a combination of depositional and diage-netic features, drives the kinetics of the thermal decompositionof the selected limestones as well as it appears to influence thetemperature of crystallisation of the main clinker phases andthe uncombined CaO content in the final clinker. In particular,the limestone with the lowest micrite to sparite ratio (1)exhibits the lowest Apparent Activation Energy (Ea)value and the highest rate of calcination and (2) requiresa lower temperature for observing the clinker phasescrystallisation and has the lowest content of uncombined
CaO in the final clinker, thus reflecting a high burnability ofthe limestone.
Introduction
Portland cement is usually produced by heating a mixture oflimestone and clay minerals at ~1450 °C. The main reactionstaking place during this burning process are usually reportedaccording to the temperature range at which they occur (Tay-lor 1990). The term clinkering refers to the reactions at 1300–1450 °C where a melt is formed and the materials start tonodulise. At temperatures below 1350 °C the most importantreactions are the decomposition of clay and carbonate min-erals. This latter process, commonly known as calcining orcalcination, is one of the most energy consuming reactionssince decarbonation is highly endothermic. In the following,the term ‘calcination’ is always referred to the process of ther-mal decomposition of carbonates.
In modern and efficient kilns (dry process kiln), cycloneexchanger and precalciner systems were introduced forobtaining the whole decarbonation of the raw meal beforeentering in the rotary kiln (burning zone) and 60 % of the totalfuel required by the kiln system is introduced into theprecalciner (Moffat and Walmsley 2005). The industrial pro-cess of cement production requires a huge quantity of lime-stone since it provides the source of calcium oxide: in partic-ular about 70 to 80 wt.% of the primary raw materials consistsof limestone. Note that limestone is the general term used for awide variety of carbonate sedimentary rocks made up chieflyof calcium carbonate.
As a consequence, a right selection of the starting rawmaterials and, in particular, of the limestone, is of paramountimportance as it results in more efficient cement manufactur-ing methods and it reduces the considerable environmental
Editorial handling: H. Poellmann
* Nicoletta [email protected]
1 Dipartimento di Scienze della Terra BArdito Desio^, Università degliStudi di Milano, Via Botticelli 23, 20133 Milan, Italy
2 CTG- Italcementi Group, Via Camozzi 124, 24121 Bergamo, Italy
Miner PetrolDOI 10.1007/s00710-015-0398-y
impact of cement production in terms of energy reduction andefficiency opportunities (i.e., reducing the high costs of fuelinvolved).
The important chemical process involving decarbonation ishardly influenced by the limestone physico-chemical features(Lech 2006a, b; Beruto et al. 2010). In particular, several stud-ies pointed out that the differences in lime reactivity duringquicklime manufacturing are attributed to a variation in themicrostructure of the calcined limestone (Soltan et al. 2011,2012; Alaabed et al. 2014). Moreover, in a recent study weobserve that the mineralogical composition, in particular thepresence of dolomite in the limestone, may speed up the cal-cite decomposition, thus reducing the heat and the time re-quired for its wholly carbonation (Marinoni et al. 2012).
The thermal decomposition of carbonates has been studiedin detail due to its importance in petrological processes as wellas for its considerable practical interest (constructions, steelmanufacture, chemical process, etc.; cf. Boynton 1980;Kerrick and Connolly 2001; Escardino et al. 2010). At pres-ent, the role of limestone properties (i.e., the types and quan-tities of impurities, degree of calcite crystallinity, microstruc-ture) on the dissociation reaction is not well understood (Lech2006c; Moropoluoua et al. 2011).
In such a light, a detailed insight into the kinetics of de-composition of three different quarry limestones that usuallysupply materials for cement manufacturing plant was per-formed. This research aims to evaluate the influence of lime-stone petrographic characteristics (i.e., texture, micrite/spariteratio, grain composition and crystal size) on the carbonatedecomposition in terms of mechanism and kinetics of disso-ciation. In particular, a preliminary characterisation of thelimestone samples was performed by X-Ray Powder Diffrac-tion (XRPD) and Optical Microscopy (OM). Then, isothermaltreatments for following the calcite decomposition was per-formed by means of in situ High Temperature X-Ray PowderDiffraction (HT-XRPD) analyses and Scanning Electron Mi-croscopy (SEM). Moreover, to investigate the influence of thelimestone features on the reactivity of the cement raw mix andagglomeration process in a rotary kiln, a burnability test wasperformed.
2. Sampling
In the present study, three different limestones were consid-ered (Fig. 1a). The first, hereafter labelled as RE, was sampledfrom one of the most important quarry used to produce cementin Northern Italy (Lombardy region) and belongs to theBFormazione della Corna^ of Jurassic age (Boni andCassinis 1973) (Fig. 1b). The second limestone, hereafter la-belled as GR, Cretaceous in age, was quarried in Central Italy(Molise region) and belongs to the BFacies Abruzzesi^(Sartori and Crescenti 1962) whereas the third limestone
sample comes from Southern Italy (Sicily region) and belongsto the Jurassic BUnità Panormidi^ and is hereafter addressedas IDF (Abate et al. 1978).
The quarried limestone samples were comminuted in a jawcrusher in order to reach the grain size corresponding to thesize fraction commonly required for the raw meal during clin-ker production. In particular, the raw meal is ground to about15 % residue on 88 μm sieve and correspondingly to 1.5–2.5 % residue on 212 μm (Chatterjee 2011).
Experimental
Limestone preliminary characterisation
The samples were examined using a petrographic microscopeby means of a Carl Zeiss Jena microscope in transmitted planepolarised light (PL) and crossed polarised light (XN). Photo-micrographs were acquired using a Nikon Coolpix 990 digitalcamera at 2272×1704 resolution. Thin sections were pointcounted for a semi-quantitative analysis of microfacies com-ponents (carbonate rock texture, matrix and grain types).Limestone texture was described according to the classifica-tion of carbonate rocks proposed by Dunham (1962) andEmbry and Klovan (1971).
The chemical compositions of the rawmaterials as given inTable 1 were determined by X-ray fluorescence (XRF) by aPhilips PW-2400 WD-XRF instrument.
An accurate determination of the phase content within thesamples was performed by means of X-ray Powder Diffrac-tion (XRPD) analysis. In particular the samples have beenback-loaded on a flat sample-holder and measured by aBragg-Brentano geometry PANalytical X’Pert Pro Diffrac-tometer, using CuKα radiation (1.5417Ǻ, 40 kV and40 mA), over the angular 2θ-range 5–80°, with a divergenceslit of 1/2° as instrumental setting with a counting time of 30 s/step and with a 0.02° step. A XRPD qualitative analysis hasfirst been performed by means of the PANalytical X’PertHighScore software. The phase contents have been obtainedtreating XRPD patterns by the full profile Rietveld Method(Young 1993) implemented in the GSAS software withEXPGUI as the graphical interface (Toby 2001).
Kinetic analysis of limestone
Experimental
A preliminary in situ High-Temperature X-ray Powder Dif-fraction (HT-XRPD) analysis was performed with the aim ofdefining the starting temperature for limestone dissociation onone hand, and for defining the most appropriate temperaturesfor following the limestone thermal decomposition by isother-mal runs, on the other. In particular, data were collected by
N. Marinoni et al.
means of a Philips X’Pert diffractometer, in θ-θ Bragg-Brentano geometry, equipped with a furnace (AHT-PAP1600). The XRPD patterns were recorded from roomtemperature (RT) up to 700 °C with a heating rate of25 °C/min, and from 700 to 900 °C every 10 °C, with aheating rate of 5 °C/min. The experiments were carried outusing Cu radiation, in the 2θ-angular range of 20–35°, with astep size of 0.05° 2θ and a counting time of 1 s/step.
Then, the limestone thermal decomposition was performedin isothermal condition by means of in situ High-TemperatureX-ray Powder Diffractometer (HT-XRPD), that is by monitor-ing the time evolution of some diffraction peaks at given tem-peratures. In particular the diffraction peak (104) for calcitewas chosen for following limestone dissociation as a functionof time at given temperatures. The data collections were per-formed on the angular range of 28.5–39° 2θ, with a 2θ stepsize of 0.03° and counting time of 1 s/step.
Data treatment of kinetic results
The isothermal treatments for the kinetics of the calcite wasfollowed by measuring the integrated intensities of the Braggpeaks. The peak intensity was first normalized with respect toits intensity observed in the sample at completion of the solid
state reaction involved, and then used to calculate the fraction-al conversion value, the so-called α-parameter. The idea be-hind the isothermal kinetics analysis is, as always, to try tolinearize the α vs. time data by fitting them to a model, inorder to get the value of the reaction rate constant k for eachisotherm temperature. The test models were taken from theliterature (Table 1 of Khawam and Flanagan 2006); the mostappropriate mechanistic model was chosen, based on thehighest correlation coefficient of the regression curve.
In more detail, the first-order (F1) and second-order (F2)equation models (Khawam and Flanagan 2006) appear to bethe most appropriate mechanism for describing the thermaldissociation of calcite. From the dependence of the reactionrate constants k with temperature, based on the Arrheniusequation, the values of the apparent activation energy (Ea)were evaluated. One has to bear in mind that the activationenergies obtained with this method are only apparent, as theyare obtained without taking into account, for example, thediffusion coefficients of the various chemical species, or therate of removal of the newly formed carbon dioxide, they maystill be useful to be compared with each other in this particularcase. We are dealing, in fact, with limestones different from apetrographic point of view, all treated in the same way, andheated exactly in the same environment. The activation
Fig. 1 a Sampling area of thethree limestones; b detail of thequarry where RE limestone wassampled. The symbol blacksquare refers to the sampling areof RE, GR and IDF, respectively
Table 1 Chemical composition (%) of the raw materials for the production of Portland cement clinker
Sample Chemical composition (%) by XRF
SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O L.o.I. Sum
Limestone RE 0.39 0.12 0.04 52.7 2.35 <0.006 0.16 <0.04 44.08 99.84
IDF 3.61 1.04 0.39 53.6 0.8 0.5 0.24 0.17 39.3 99.65
GR 0.35 0.09 0.04 54.65 0.43 0.02 0.28 0.01 43.98 99.81
Clays 43.99 23.21 9.27 0.28 1.1 0.04 – 1.68 19.14 98.71
Si oxide 99.10 0.07 0.01 0.01 – – 0.02 0.03 0.70 99.87
Fe oxide 0.1 0.10 99.50 – – – – – 0.65 100.35
The role of petrography on the thermal decomposition
energies obtained are obviously valid only in the conditionswe used, but they are still useful to compare the performancesof the different limestone under study.
Burnability test
This test involves raw materials in a laboratory furnace up toclinkering temperature for a period similar to the residencetime in the kiln and then testing the resultant clinker for thedetermination of the uncombined CaO (free lime).
In the present study, each limestone was mixed in differentpercentage with Si and Fe oxides and clay minerals, respec-tively (Table 2), in order to obtain three laboratory raw mixeswith a chemical composition fitting the typical kiln feed mix.The raw mixes were thermally treated at 1000, 1100, 1200,1300, 1350, 1400 and 1450 °C respectively in an electricalfurnace with a soaking time of 1 h and then rapidly cooled inair with the aim of simulating the sintering and cooling con-ditions in an industrial kiln. The uncombined CaO was deter-mined by using the Franke method (Pressler and Brunauer1956). Moreover, the thermally treated samples were also ex-amined by means of XRPD and Scanning Electron Microsco-py (SEM) in order to assess the mineralogical compositionand the texture of the sintered materials.
4. Results
Limestone preliminary characterisation
Mineralogically, the selected limestones are chiefly composedby calcite associated with minor contents of quartz (Table 3);dolomite only occurs as an minory phase in RE samples. Pet-rographically, the limestone samples display a variety of tex-tural features and grain composition as reported in Table 4.
The RE limestones (Table 4; Fig. 2a–f) include peloidalskeletal wackestone (microfacies A; Fig. 2a), peloidal skeletalgrainstone to packstone (microfacies B; Fig. 2b), peloidalcoated grain grainstone (microfacies C; Fig. 2c), and calci-mudstone with partial dolomitization (microfacies D;Fig. 2d–f). Wackestone and calci-mudstone are matrix-supported textures (microfacies A and D) consisting of micro-crystalline calcite (micrite), which locally underwent
diagenetic recrystallization to microsparite owing toaggrading neomorphism. Microfacies A wackestone includessparse peloids and skeletal fragments (benthic foraminifers,bivalves, algae and ostracodes). The grainstone to packstonetextures of microfacies B are compacted with concave-convexgrain contacts and minor post-compaction sparite cementa-tion. Microfacies C grainstone contains peloids, cortoids,and skeletal fragments, such as benthic foraminifers, bivalvesand echinoderms. Cement filling interparticle space consistsof isopachous rims ofmarine fibrous cement followed by clearblocky mosaics of equant sparite, precipitated in a burial dia-genetic environment. In microfacies D, irregular patches ofdolomite consists of euhedral rhombohedral crystals replacingthe micrite matrix, attributable to secondary dolomitisationprocesses. In microfacies A and D, the carbonate mud matrixis cross-cut by a few hundredmicrons tomillimetre wide veinsfilled by blocky sparite; these veins penetrate across the sed-imentary structure and the filling cements were likely precip-itated by meteoric waters or burial brines that percolatedthrough fractures. Chert particles, produced by late diageneticsilicification of the limestone, are occasionally present in thethin sections.
The IDF limestones (Figs. 2g–h and 3a–e) comprisepeloidal wackestone to calci-mudstone with rare fenestrae(microfacies A; Fig. 2g–h), peloidal skeletal packstone towackestone (microfacies B; Fig. 3a), pisoidal peloidalgrainstone with rare ooids (microfacies C; Fig. 3b–c), calci-mudstone with rare fenestrae (microfacies D; Fig. 3d), anddasycladacean algae wackestone (microfacies E; Fig. 3e). Inmicrofacies A and D, the limestone texture consists of micritematrix embedding sparse peloids and rare skeletal fragments,characterized by syn-sedimentary millimetre-size cavitiesgeopetally filled by sediment at the base and blocky sparitecement at the top. This voids are labelled as fenestrae, andattributed to deposition in intertidal environment. MicrofaciesB wackestone to packstone consists of peloids of variable size,bivalve fragments and benthic foraminifers. Grainstones(microfacies C) include pisoids, cortoids, aggregate grains,bivalve shells with micrite envelopes, benthic foraminifersand minor ooids. Cement in grainstones consists of a firstgeneration of marine isopachous rims of fibrous cementaround the grains, followed by burial equant calcite sparite.In some samples, meniscus micrite coatings of vadose
Table 2 Amount in weightpercentage (wt%) of limestones,clays and oxides used forpreparing the initial feed mixes
Raw mix Limestones (wt%) Clays (wt%) Si oxide (wt%) Fe oxide (wt%)
RE GR IDF
RE_mix 78.11 – – 13.64 7.37 0.89
IDF_mix – – 81.25 11.22 6.58 0.95
GR_mix – 77.40 – 14.13 7.56 0.90
N. Marinoni et al.
meteoric diagenesis bind the grains. Microfacies E is rare andconsists of wackestone with fragments of the greendasycladacean alga Clypeina jurassica.
The GR limestone microfacies (Fig. 3f–h) consists ofcompacted bioclastic packstone to rudstone, medium to coarsesand to gravel grain size. Skeletal fragments include dominantrudist bivalves, echinoid spines and crinoid ossicles, sparsefragments of benthic foraminifers and peloids. Grain contactsare often sutured due to compaction and pressure solution.Micrite matrix is locally present in between the compactedbioclastic grains. Minor syntaxial calcite cement precipitatedaround echinoderm fragments. The GR limestone has an ap-parently coarse crystalline appearance at the mesoscale obser-vation due to the specific nature of the carbonate shell struc-ture of the skeletal components. Crinoid ossicles and echinoidplates and spines form as single calcite crystals, whereasrudists were characterized by up to several centimetere sizeand several millimetres thick bimineralic (calcite and arago-nite) shells.
All the investigated limestones did not show identifiableporosity at the optical microscopy scale of investigation.Whether microporosity (<1 μm pore size) within the micritematrix is present, it could not be identified through petro-graphic analysis.
Isothermal kinetic analysis
In the present section the results from the isothermal analysesby means of in situ HT-XRPD and the SEM observations onthe heated samples are reported.
Figure 4 displays the isothermal α-t curves of calcite de-composition in the samples. The α vs. T shows a similar linearsteady-isothermal kinetics until approximately 95 % of thecalcite has decomposed. This hints that calcite decompositionfollows a mono-step process, which implies one only apparentactivation energy value. In the samples IDF and RE, the cal-cite decomposition is well described by a linear reaction mod-el; conversely, in GR a second-order model is suggested(Fig. 5). In Table 5 the Ea’s values are reported.
In Table 5 the time to achieve an α value of 0.5 is reported.RE and IDF exhibit t (α=0.5) of ~ 55 and ~ 45 min
respectively. In the case of GR, T should be higher than600 °C for starting the calcite decomposition. Therefore the t(α=0.5) of Gr is not comparable with the others.
As a whole, in sample RE the presence of dolomite speedsup the calcite decomposition in terms of Ea and t (α=0.5)values as previously observed by Marinoni et al. 2012.
After isothermal runs all the heated samples reveal a similarmicrostructure: they are chiefly composed of crystals of limewhere the crystals have been growing conserving the originalcalcite shape (Fig. 6a). Fissures develop in the CaO micro-crystals, thus facilitating the CO2 diffusion from calcium car-bonate decomposition through the very porous CaO structureformed in each microcrystal (Fig. 6b). (cf. similar results inLech 2006b).
Burnability test
A similar trend in terms of phase occurrence is observed in allthe investigated samples. In particular, XRD patterns from1000 to 1450 °C provide an overview of the main reactionstaking place upon heating: the decomposition of clay mineralsand calcite occurs to give gehlenite with free lime followed bythe belite formation associated with aluminates and ferrite.Only when T reaches 1350 °C the Bragg peaks belonging toalite appear.
The investigated samples differ from each other in thecrystallisation T of the main clinker phases. For instance, inRE_mix ferrite and aluminate as well as belite occur at1100 °C whereas a T of 1200 °C is required for observingtheir growth in IDF_ and GR_mix, respectively (Table 6).
Once more, the trend of the free lime content as a functionof the treatment temperature highlights the lowest free limecontent at all the temperatures in RE_mix (Fig. 7). TheGR_mix has the highest free lime content up to 1200 °C, thena rapid decreasing is observed between 1200 and 1350 °C,then reaching the same free lime contents as RE_mix. TheIDF_mix presents a hard burnability with high values of freelime also the highest T explored (~10 wt% at 1450 °C).
All samples heated at 1450 °C consisted of the typicalclinker phases (Fig. 6c): high amounts of alite grains andC2S crystals or cluster are surrounded by C3A and C4AF
Table 3 Mineralogical composition and micrite to sparite ratio (M/S) of the selected limestones
Sample Mineralogical phases (wt%) Limestone Textural property
Calcite Dolomite Quartz Clays* Mica* Micrite/Sparite ratio (M/S)
RE 90.3(1) 7.5(1) 2.2(1) < 1 < 1 2.2
IDF 98.6(1) – 1.4(1) < 1 < 1 1.5
GR 98.7(1) – 1.3(1) < 1 < 1 0.5
Symbol: * refers to the mineralogical phases detected after chemical attack with dilute HCl on the bulk sample. The values in brackets refer to thestandard deviation value
The role of petrography on the thermal decomposition
Tab
le4
Texturalfeatures
andgraincompositio
nof
theinvestigated
limestone
samples
Location/Unit/
Age
Microfacies
Texture
Com
ponents(grain
size)
Diagenetic
features
Depositionalenvironm
ent
Rezzato
(RE)/
Formazione
della
Corna
Jurassic
RE m
icrofacies
A(Fig.2a)
Peloidalcalci-
mudstone,
wackestoneandrarely
packstonewith
rare
skeletal
fragments
Com
mon
peloids(sizefrom
50to
200μm),somefaecal
pellets(upto
500μm),rare
cortoids,rareskeletal
fragmentssuch
asalga
Thaumatoporella
parvovesiculifera,
calcispheres,textularids
andother
smallb
enthicforaminifers,ostracodes,bivalves
Recrystallizationof
micritein
microsparite
(5–10μm;
aggradingneom
orphism);
isopachous
fibrouscement
with
inintraparticlepore
inside
algae;burialblocky
sparite
with
inundeterm
ined
biom
olds;tectonic
fracturesandveinswith
blocky
sparite
Shallowcarbonate
platform
interior,
restricted
subtidallow-
energy
lagoon
RE m
icrofacies
B(Fig.2b)
Com
pacted
peloidal
skeletalgrainstone
topackstone
Abundantp
eloids
(grain
size
100–500μm),benthicforaminifers
Com
paction:
elongatedand
concave-convex
grain
contacts;rarepost-com
paction
cementatio
n:10–30μm
thick
rim
ofequant
microsparite/
sparite
cement
Shallowcarbonate
platform
interior,
subtidalmoderate
energy
areasin
lagoon
RE m
icrofacies
C(Fig.2c)
Peloidalcoatedgrain
(rareooidsand
cortoids)grainstone
Com
mon
peloids(upto
500μm),rare
ooids(peloids
atnucleus),cortoids(m
icritecoated
bioclastswith
micriteenvelopes),benthicforaminifers,fragments
ofcalcifiedmicrobessuch
asCayeuxia
Isopachous
rimsof
fibrous
cement(marinediagenesis)
Blockysparite
mosaics
(buriald
iagenesis)
Openmarine,shallow,
abovewavebase,high
energy
facies
ofbeach
orshoals
RE m
icrofacies
D (Fig.2d–f)
Calci-m
udstonewith
sparse
peloidsand
dolomite
crystals
Rare/sparse
peloids,andundeterm
ined
biom
olds
filled
bycalcite
blocky
sparite;d
etritalsand-gradequartzgrains
Sparsesilicifiedareas
Recrystallizationof
micritein
microsparite;b
locky
sparite
inbiom
olds
(burial
diagenesis);euhedraldolomite
crystals(100–300
μm),with
dedolomitizatio
n;silicification;
veinsfilledby
blocky
sparite
(from
100μm
toseveralm
ms
wide)
Shallowcarbonate
platform
interior,
restricted
low-energy
subtidalto
intertidal
Isoladelle
Fem
mine
(IDF)/Unità
Panormidi
Upper
Jurassic
Kim
meridgi-
an-Tith
onian
IDF microfacies
A (Fig.2g–h)
Peloidalw
ackestoneto
calci-mudstonewith
rare
fenestrae
Peloids(sizefrom
50to
400μm),
rare
textularid
foraminifers,ostracodes,undetermined
biom
olds
filledby
blocky
sparite;irregular
vugs
(fenestrae)
filledgeopetally
bypeloidalwackestoneandblocky
sparite
Recrystallizationof
micritein
microsparite;fenestrae,
rare
replaced
biom
olds
and
fracturesfilledby
burial
blocky
sparite
Shallo
wplatform
interior,
low-energyrestricted
subtidalto
intertidal
IDF microfacies
B(Fig.3a)
Peloidalw
ackestoneto
packstonewith
sparse
skeletal
fragments
Peloids(sizefrom
50to
500μm),rare
benthic
foraminifers(textularids),bivalvefragmentswith
micrite
envelopes,very
rare
pisoidsandcoated
calci-
mudstoneintraclasts(upto
1mm
insize)
Recrystallizationof
micritein
microsparite;fenestrae,
replaced
bivalveshellsand
fracturesfilledby
burial
blocky
sparite
Shallo
wplatform
interior,
lowenergy
from
intertidalto
restricted
subtidallagoon
IDF microfacies
Pisoidalp
eloidal
grainstone
toPeloids(100–500
μm),pisoids(400
μm
to1mm),
aggregategrains
(upto
1mm),bivalves
with
micrite
Isopachous
rimsof
fibrouscement
(marinediagenesis);
Shallo
wplatform
from
intertidal
N. Marinoni et al.
Tab
le4
(contin
ued)
Location/Unit/
Age
Microfacies
Texture
Com
ponents(grain
size)
Diagenetic
features
Depositionalenvironm
ent
C (Fig.3b–c)
packstonewith
rare
ooids
envelopes,broken
ooids,fragmentsof
calcified
filamentous
microbessuch
asCayeuxia,fenestrae
meniscusmicritecement
(meteoricvadose
diagenesis);burialblocky
sparite
filling
interparticle
pores,fenestraeandfractures
(buriald
iagenesisor
meteoricphreatic)
tosupratidalhigh-
energy
beachor
shoal
IDF microfacies
D(Fig.3d)
Calci-m
udstonewith
rare/sparsepeloids
with
rare
fenestrae
Peloids
(50–400μm)
Recrystallizationof
micritein
microsparite;fenestrae
andfracturesfilledby
blocky
sparite
(burial
diagenesisor
meteoricphreatic);
subaerialexposure
andcalcretepaleosol
developm
entw
ithcircum
granular
cracks
Shallo
wplatform
interior,
low-energyintertidal
flat,
periodically
exposedto
subaerialconditio
ns
IDF microfacies
E(Fig.3e)
Skeletalp
eloidal
wackestonewith
dasycladaceanalgae
Fragmentsof
dasycladaceanalgaeClypeinajurassica,
peloids,bivalvebiom
olds
Recrystallizationof
micritein
microsparite;
undeterm
ined
biom
olds,
bivalves
andfracturesfilled
byblocky
sparite
Shallowcarbonate
platform
interior,subtid
allow-
energy
lagoon
Guardiaregia
(GR)/Facies
Abruzzesi
Upper
Cretaceous
GR m
icrofacies
(Fig.3f–h)
Bioclastic
packstoneto
rudstone
with
rudist
andechinoderm
fragments
Dom
inantfragm
entsof
rudistbivalves
(0.5–2
mm),echinoderm
s(upto
1mm)includingechinoidsspines,platesandcrinoidossicles,
fragmentsof
benthicforaminiferssuch
asOrbito
ides
andSiderolites,
rare
peloids,bryozoan
fragments,tubular
calcim
icrobes
Com
pactionwith
suturedgrain
contacts;m
inor
syntaxialcem
ent
around
echinoderm
sFragm
entsof
karstic
cave
speleothem
ecement
Carbonaterampor
open
marine
platform
reworking
rudis t
bioherms
The role of petrography on the thermal decomposition
which formed the clinker melt phase during heat treatment.Widely dispersed periclase is observed in RE_mix as well as
cluster of rounded grains of lime are clearly visible inIDF_mix (Fig. 6d). The clinker produced with IDF_mix
Fig. 2 Photomicrographs of the investigated RE limestone samples. a RElimestone microfacies A peloidal skeletal wackestone with peloids (P)and texturalrid foraminifers (T) cross-cut by a fracture filled by blockysparite. b Microfacies B of RE limestone showing compacted grainstonewith peloids (P) and benthic foraminifers (F). c RE microfacies Cgrainstone with grain-supported texture of peloids and cortoids (micritecoated grains) bound by marine fibrous cement (C). d RE microfacies Dconsisting of calci-mudstone with sparse dolomite rhombic crystals (D).The thin section is stained with alizarine red, which colours in pink onlycalcite leaving the dolomite crystals unstained. Alizarine staining allowsidentifying evidences of de-dolomitization because the dolomite crystals
have patches of calcite overgrowth identified by pink colour areas on therhombic dolomite crystals. eMicrofacies D ofRE limestone with partiallydolomitized (D) calci-mudstone. The grain of the left (S) consists of chertdue to diagenetic limestone silicification. f Same image as in (e) incrossed polarizers (XN) showing the dolomite crystals (D) and the chertgrain on the left side (S). g IDF limestone microfacies A peloidalwackestone with rare skeletal fragments such as foraminifers (F). h IDFlimestone microfacies A peloidal wackestone with ostracodes (O) andfenestral void (F) filled geopetally with micritic sediment at the baseand blocky sparite cement roofing the syn-sedimentary cavity. A verticalfracture cross-cuts the calci-mudstone and fenestrae
N. Marinoni et al.
featured a high macro-porosity in the range of 800 μm, thusreflecting a hard burnability. In contrast, a very compact mi-crostructure is observed in the clinker formed by usingRE_mix.
Discussion
The XRD characterisation shows that the three limestonesappear similar from a mineralogical point of view but exhibitinhomogeneities in terms of texture as pointed out by
petrographic observations of thin sections. In particular, REis chiefly composed by calci-mudstone and wackestone asso-ciated with minor grainstone, suggesting a Jurassic platforminterior with shallow low-energy restricted depositional envi-ronments (subtidal lagoon, intertidal flat) for the former, andhigh-energy above wave base for the latter. The IDF Jurassiclimestone grainstone microfacies is interpreted as representinghigh-energy subtidal to supratidal shoal or shoreline depositsadjacent to restricted subtidal lagoon and tidal flat with calci-mudstone and wackestone with fenestral voids and events ofsubaerial exposure. Conversely, the Upper Creatceous GR
Fig. 3 Photomicrographs of theinvestigated IDF and GRlimestone samples. a IDFmicrofacies B peloidal skeletalwackestone to packstone withpeloids (P), bivalve fragments (B)with micrite envelope andtextularid foraminifers (F). b IDFmicrofacies C grainstone withpeloids (P) and pisoids (P) andmarine fibrous cement. c Crossedpolarizers (XN) photomicrographof IDF microfacies C grainstonewith grains bounded by a rim ofisopachous fibrous cement ofmarine diagensis followed byburial blocky sparite. d IDFmicrofacies D calci-mudstonecross-cut by cement-filledfractures. e IDF microfacies Ewackestone with the greendasycladacean alga Clypeinajurassica. f GR microfaciespackstone to rudstone withmillimetre size crinoid ossicle(C), which consists of a singlecrystal of calcite, associated withformanifer and rudist shellbioclastic debris. g Parallelpolarizers and h crossedpolarizers image of GRcompacted skeletal packstonewith fragments of benthicforaminifers (F), crinoids andrudist bioclastic debris
The role of petrography on the thermal decomposition
limestone represents open marine, bioclastic packstone torudstone accumulated in a high energy, wave and currentdominated carbonate ramp reworking rudist and echinodermfragments. RE and IDF exhibit a micrite/sparite ratio of 2.2and 1.5 respectively, whereas the lowest one is observed inGR(M/S=0.5).
Before discussing the kinetic results (i.e.,Ea values and rateof calcination) one has to bear in mind that our results areobviously valid only in the experimental set-up we used. Inthe present research, the limestone decomposition is per-formed under isothermal conditions and its dissociation wasfollowed in a static atmosphere of air where the composition
Fig. 4 Calcite decomposition followed by means of in situ HT-XRPD:calcite α-t curves for the three isothermal runs (°C) as a function of time
Fig. 5 Lnln plot of the isothermal runs vs. ln t
Table 5 Calcite Kinetics: time (min) to achieve an α-value of 0.5 aswell as Apparent Activation Energy (Ea) and correlation factor (R
2) of thelinear regression analysis in the Arrhenius plot
Sample Calcite kinetics
t(min) α=0.5 at 600 °C Ea (kJ/mol) R2
RE 45 834 0.982.2
IDF 55 1229 0.991.5
GR 100* 1554 0.980.3
*It refers to an α-value calculated at 640 °C, therefore not comparablewith the others
N. Marinoni et al.
of the local atmosphere within the hot chamber changes inresponse to the liberation of CO2 as the carbonate decompo-sition proceeds. Therefore, our kinetic data should not be con-sidered as absolute, but always related to the actual conditionsof the experiments. However, a direct comparison between thekinetic results for the three samples can be certainly drawn.
As a whole, the limestone decomposition is characterisedby the dissociation of calcite into CaO with release of CO2,thus leading to a general microcracking of the heated samples;only in RE the CaCO3 decomposition follows the one of do-lomite which takes place in ~ few minutes, leading to CaOassociated withMgOwith CO2 release. The limestone decom-position is accompained by the presence of diffusemicrocracks in the calcined samples during isothermal treat-ments. This kind of cracks are labelled as contraction cracks,which develop during calcination due to contraction of thesample linear sizes caused by thermal decomposition of calci-um carbonate (Lech 2006b). In contrast, no evidences of graincracks that lead to a crumbling of the samples into grains as
well as of interface cracks that occur between neighbouringphases with different molar volumes are observed.
The inhomogeneities among the samples seem to influ-ence, on the one hand, the kinetics of calcite decompositionand, on the other, the burnability behaviour of the limestone.
RE and IDF, comparable from a petrographic point of view,age and environment of deposition, exhibit the same graphicalshape of their isothermal runs (α vs. t), suggesting that thecalcite decomposition is well described by the first-orderequation (F1). For GR, where a grain-supported texture pre-vails with grains consisting of calcite crystals due to the natureof the dominant skeletal material such as echinoderms and
Fig. 6 SEM images: a CaOcrystals preserving the originalrhombohedral calcite shape (REafter isothermal run at 610 °C); bmicrocracks (highlighted withwhite colour) within lime crystalsgrowing during isothermaltreatment (IDF after isothermalrun at 610 °C); c coarse grainedclinker with uniform phasedistribution (RE_mix heated at1450 °C); dmacro-porosity in theclinker obtained after burning testat 1450 °C, thus reflecting a lowburnability of the IDF_mix initialfeed mix (sample IDF_mix heatedat 1450 °C)
Table 6 Temperature of crystallisation (°C) of the main clinker phasesdetermined during in-situ HT-XRPD experiments
Raw mix T of crystallization (°C) of the clinker phases
C3S C2S C3A C4AF
RE_mix 1300 1100 1100 1100
IDF_mix 1400 1200 1200 1200
GR_mix 1400 1200 1200 1200 Fig. 7 Weight percentage (wt%) of CaO calculated in the clinker heatedat different T during burning test
The role of petrography on the thermal decomposition
rudist bivalves, a model corresponding to second-order reac-tion (F2) is required.
The limestone texture and grain type composition (depend-ing on depositional environment and age of the carbonaterock) seem to be the controlling features on the kinetics ofcalcination (i.e., T and rate of calcination) leading to differentapparent activation energy values that range from 834 to1559 kJ/mol. Apparently, the Ea-values show a monotonicallydecreasing dependence on the M/S value: the higher themicrite/sparite ratio values, the lower the magnitude of Ea.Note that the lowest value of Ea for RE can also be relatedto the presence of dolomite as accessory phase in the lime-stone, as previously observed by Marinoni et al. (2012). Inparticular, the limestone dissociation begins with the dolomitedecomposition; the resulting cracks due to the CaMg(CO3)2dissociation allow an easier way of CO2 diffusion through theproduct layer, increasing the sample surface area where thecalcite dissociation can occur, thus requiring low energy forthe CaCO3 decomposition.
Once more, a linear correlation is found between the M/Sratio and the time to achieve the same α conversion value at agiven T, suggesting a different rate of calcination among thesamples: the higher the micrite-sparite ratio, the lower the timeof performance in the hot chamber for reaching the wholecalcite decomposition.
Finally, the high content of free lime even at high T inIDF_mix as well as the high macro-porosity reported forGR_mix reflect the difficulty in the transformation of the rawmaterial into the desired clinker phases. This correlates wellwith the high heat input required for starting the calcite de-composition in GR (i.e., high Ea value) as well as a relation-ship is found between the low rate of calcination observed inIDF with the hard burnability of IDF_mix, thus leading highlevels of CaO up to 1450 °C. This implies a longer residencetime into a rotary kiln with the goal of obtaining similar clin-ker phase composition as in the industrial scale. On the con-trary, RE_mix facilitate the achievement of the burnabilitygoal as pointed out by the loss of free lime associated withthe growth of alite even at 1350 °C, thus confirming the lowheat input (i.e., low Ea value) and high rate of calcinationreported by kinetics data for the CaCO3 decomposition in RE.
6. Conclusion
A study on the thermal decomposition of three cement-grade run-on-mine limestones has been performed bymeans of isothermal analysis. In particular, the influenceof the limestone texture and carbonate grain composition(determined through petrographic analysis) on the kineticsof carbonate dissociation as well as on their burnabilitybehaviour was taken into account.
The limestones were quarried and further comminuted tothe standard grain size usually required for the raw meal com-monly injected into the precalciner-preheater system.
The kinetics results attest that limestone texture coupledwith the micrite/sparite ratio become interdependent and in-terrelated factors highly affecting the limestone T of decom-position as well as its kinetics, the latter in terms of Ea valuesand the rate of dissociation.
The present results have fundamental implications at indus-trial scale during the cement manufacturing. From the experi-ence gained with different types of precalciner kilns thereappears to be a relationship between the variations of degreeof calcination and the residence time of the material in theprecalciner as well as a narrow particle size distribution isoptimally required as fines tend to increase dust loss by en-trainment in exhaust gases, while coarse particles are harder toreact in the kiln. On the contrary we observed that the textureheavily influence the T for starting the carbonate decomposi-tion as well as its rate of decomposition. In particular, thehighest reactivity is observed for the raw meal where themicrite/sparite ratio is the highest, thus enhancing the calcitedissociation in terms of energy input (T) and time for achiev-ing the complete calcination, thus reducing the heat and ener-gy consumption during the decomposition processes.
Acknowledgments We are grateful to Dr. Emanuele Gotti and FedericaAngeloni from CTG for their support during SEM investigations andMarco Cantaluppi for his help during the XRPD analysis.
References
Abate B, Catalano R, Renda P (1978) Schema geologico dei Monti diPalermo (Sicilia). Boll Soc Geol Ital 97:807–819
Alaabed S, Soltan MA, Abdelghany O, Amin BEM, Tokhi ME, KhaleelA, Musalim A (2014) United Arab Emirates limestones: impact ofpetrography on thermal behaviour. Mineral Petrol. doi:10.1007/s0710-014-0329-3
Beruto DT, Botter R, Cabella R, Lagazzo A (2010) A consecutivedecomposition-sintering dilatometer method to study the effect oflimestone impurities on lime microstructure and its water reactivity.J Eur Ceram Soc 30:1277–1286
Boni A, Cassinis G (1973) Carta geologica delle Prealpi Bresciane a Suddell’Adamello. Atti Ist Geol Univ Pavia 23:119–159
Boynton RS (1980) Chemistry and technology of lime and limestone.Wiley Interscience, New York
Chatterjee AK (2011) Chemistry and engineering of the clinkerizationprocess — Incremental advances and lack of breakthroughs. CemConcr Res 41:624–641
Dunham RJ (1962) Classification of carbonate rocks according to depo-sitional texture. AAPG Mem 1:108–121
Embry AF, Klovan JE (1971) A Late Devonian reef tract on northeasternBanks Island. Bull Can Petrol Geol 19:730–781
Escardino A, García-Ten J, Feliu C, Moreno A (2010) Calciumcarbonate decomposition in white-body tail during firing. JEur Ceram Soc 30:1989–2001
N. Marinoni et al.
Kerrick DM, Connolly JAD (2001) Metamorphic devolatilization ofsubducted oceanic metabasalts: implications for seismicity, arcmagmatism and volatile recycling. Earth Planet Sci Lett 189:19–29
Khawam A, Flanagan DR (2006) Solid-state kinetic models: basics andmathematical fundamentals. J Phys Chem B 110:17315–17328
Lech R (2006a) Thermal decomposition of limestone: Part 1 – Influenceof properties on calcination time. Silic Ind 71:103–109
Lech R (2006b) Thermal decomposition of limestone: Part 2 – Influenceof contraction, phase composition, phase concentrations and heatingon calcination time. Silic Ind 71:110–114
Lech R (2006c) Thermal decomposition of limestone: Part 3 – Kineticcurves. Silic Ind 71:143–148
Marinoni N, Allevi S, Marchi M, Dapiaggi M (2012) A kinetic study ofthermal decomposition of limestone using in situ high temperaturex-ray powder diffraction. J Am Ceram Soc 95:2491–2498
Moffat W, Walmsley RW (2005) Understanding lime calcination kineticsfor energy cost reduction. Proc 59th Appita Conf, Auckland, May,487–494
Moropoluoua A, Bakolas A, Aggelakopoulou E (2011) The effects oflimestone characteristics and calcination temperature to the reactiv-ity of the quicklime. Cem Concr Res 31:633–639
Pressler EE, Brunauer S (1956) Investigation of Franke method of deter-mining free calcium hydroxide and free calcium oxide. Anorg AllgChem 28:896–902
Sartori S, Crescenti U (1962) Ricerche biostratigrafiche nel Mesozoicodell’Appennino Meridionale. Giorn Geol 50:161–168
Soltan AM, Kahl WA, Hazem M, Wendschuh M, Fischer RX (2011)Thermal microstructural changes of grain-supported limestones.Mineral Petrol 130:9–17
Soltan AM, Kahl WA, Wendschuh M, Hazem M (2012) Microstructureand reactivity of calcined mud supported limestones. Miner ProcessExtr Metall 121:5–11
Taylor HWF (1990) Cement chemistry. Academic, New YorkToby BH (2001) A graphical user interface for GSAS. J Appl Crystallogr
34:210–213Young RA (1993) The rietveld method. University Press, Oxford
The role of petrography on the thermal decomposition
本文献由“学霸图书馆-文献云下载”收集自网络,仅供学习交流使用。
学霸图书馆(www.xuebalib.com)是一个“整合众多图书馆数据库资源,
提供一站式文献检索和下载服务”的24 小时在线不限IP
图书馆。
图书馆致力于便利、促进学习与科研,提供最强文献下载服务。
图书馆导航:
图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具