Feasibility study on the use of cellular concrete as alternative raw material for Portland clinker production Joris Schoon a,b , Klaartje De Buysser c , Isabel Van Driessche c , Nele De Belie b,⇑ a S.A. Sagrex N.V., Heidelberg Cement Benelux, Heidelberg Cement Group, Terhulpsesteenweg 185, B-1170 Brussels, Belgium b Magnel Laboratory for Concrete Research, Department of Structural Engineering, Faculty of Engineering and Architecture, Ghent University, Technologiepark Zwijnaarde 904, B-9052 Ghent, Belgium c Sol gel Centre for Research on inorganic Powder and Thin films Synthesis (SCRiPTS), Department of Inorganic and Physical Chemistry, Faculty of Sciences, Ghent University, Krijgslaan 281-S3, B-9000 Ghent, Belgium highlights Cellular concrete could be used as raw material for Portland clinker production. Using cellular concrete as raw material will not have major beneficial advantages. The poor grindability of the cellular concrete will act as a serious restriction. This study reveals restrictions generalizable for other alternative raw materials. article info Article history: Received 4 April 2013 Received in revised form 10 June 2013 Accepted 21 July 2013 Available online 24 August 2013 Keywords: Clinker Cellular concrete SiO 2 Quartz Fineness 45 lm 90 lm Alite Free CaO Burnability abstract This paper aims to investigate the use of cellular concrete as an alternative raw material (ARM) for Port- land clinker kilns. The possibility to generate a raw material with a stable compositional variation was investigated as well as simulations were carried out to maximise their use in clinker kilns. Based on these simulations, experimental clinkers were produced with dosages that were esteemed as realistic. Because of the presence of important levels of quartz sand in the cellular concrete materials, the energy necessary to grind the alternative raw material in comparison with comparable classic raw materials was investi- gated as well as the influence of the final particle size distribution of the Cold Clinker Meal (CCM) on the mineralogical composition of the final clinkers. It will be demonstrated that cellular concrete materials can be used as ARM for clinker production although there are some important restrictions that will limit the practical implementation. This investigation will also provide some interesting knowledge on the use of other recycled concrete materials as alternative raw material for Portland clinker production. Ó 2013 Elsevier Ltd. All rights reserved. 1. Introduction This paper examines the possibility to use cellular concrete material as an alternative raw material (ARM) for Portland clinker production. The use of alternative raw materials which could de- crease the use of natural sources as well as counter landfill with construction waste is in line with the Cement Sustainability Initia- tive [1] described by the World Business Council for Sustainable Development. Several studies were already performed that could improve the ecological impact of Portland clinker production by using secondary or recycled materials as alternative raw material in Portland clinker production [2–8]. Cellular or aerated concrete is a type of lightweight concrete with a high concentration of air 0950-0618/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2013.07.083 Abbreviations: Ant, Antoing; LiqSimple, Liquid Simple; ACC, Autoclaved Cellular Concrete; LOI, Loss On Ignition; ARM, Alternative Raw Material; Lo, Loam (SiO 2 - source); ARM/CCC, Cellular Concrete Clean; Lxh, Lixhe; ARM/CCP, Cellular Concrete Polluted; Maa, Maastricht; CCM, Cold Clinker Meal; Ma, Marl (specific type of limestone); Cl, Clinker; PL, Poor Limestone; CRM, Classic Raw Material; Ref, Reference; Decarb E, Decarbonation Energy; RL, Rich Limestone; DoS, Degree of Sulfatisation; SC, Sabulous Clay (SiO 2 -source); FA, Fly Ash (Al 2 O 3 -source); SR, Saturation Rate; HCM, hot clinker meal; Tu, Tuffeau (specific type of limestone); IC, Iron Carrier (Fe 2 O 3 -source); XRD, X-Ray Diffraction; LSF, Lime Saturation Factor; XRF, X-Ray Fluorescence. ⇑ Corresponding author. Tel.: +32 9 264 55 22; fax: +32 9 264 58 45. E-mail address: [email protected](N. De Belie). Construction and Building Materials 48 (2013) 725–733 Contents lists available at ScienceDirect Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Transcript
Construction and Building Materials 48 (2013) 725–733
Joris Schoon a,b, Klaartje De Buysser c, Isabel Van Driessche c, Nele De Belie b,⇑a S.A. Sagrex N.V., Heidelberg Cement Benelux, Heidelberg Cement Group, Terhulpsesteenweg 185, B-1170 Brussels, Belgiumb Magnel Laboratory for Concrete Research, Department of Structural Engineering, Faculty of Engineering and Architecture, Ghent University, Technologiepark Zwijnaarde 904,B-9052 Ghent, Belgiumc Sol gel Centre for Research on inorganic Powder and Thin films Synthesis (SCRiPTS), Department of Inorganic and Physical Chemistry, Faculty of Sciences, Ghent University, Krijgslaan281-S3, B-9000 Ghent, Belgium
h i g h l i g h t s
� Cellular concrete could be used as raw material for Portland clinker production.� Using cellular concrete as raw material will not have major beneficial advantages.� The poor grindability of the cellular concrete will act as a serious restriction.� This study reveals restrictions generalizable for other alternative raw materials.
a r t i c l e i n f o
Article history:Received 4 April 2013Received in revised form 10 June 2013Accepted 21 July 2013Available online 24 August 2013
Keywords:ClinkerCellular concreteSiO2
QuartzFineness45 lm90 lmAliteFree CaOBurnability
a b s t r a c t
This paper aims to investigate the use of cellular concrete as an alternative raw material (ARM) for Port-land clinker kilns. The possibility to generate a raw material with a stable compositional variation wasinvestigated as well as simulations were carried out to maximise their use in clinker kilns. Based on thesesimulations, experimental clinkers were produced with dosages that were esteemed as realistic. Becauseof the presence of important levels of quartz sand in the cellular concrete materials, the energy necessaryto grind the alternative raw material in comparison with comparable classic raw materials was investi-gated as well as the influence of the final particle size distribution of the Cold Clinker Meal (CCM) on themineralogical composition of the final clinkers. It will be demonstrated that cellular concrete materialscan be used as ARM for clinker production although there are some important restrictions that will limitthe practical implementation. This investigation will also provide some interesting knowledge on the useof other recycled concrete materials as alternative raw material for Portland clinker production.
� 2013 Elsevier Ltd. All rights reserved.
1. Introduction
This paper examines the possibility to use cellular concretematerial as an alternative raw material (ARM) for Portland clinkerproduction. The use of alternative raw materials which could de-crease the use of natural sources as well as counter landfill withconstruction waste is in line with the Cement Sustainability Initia-tive [1] described by the World Business Council for SustainableDevelopment. Several studies were already performed that couldimprove the ecological impact of Portland clinker production byusing secondary or recycled materials as alternative raw materialin Portland clinker production [2–8]. Cellular or aerated concreteis a type of lightweight concrete with a high concentration of air
726 J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733
voids. Aggregates and cements used in standard concrete composi-tions on a daily base are found in cellular concrete. This impliesthat part of the final conclusions of this study, will also apply forregular concrete. An air-entraining agent is used to create the cel-lular structure of air voids by which the density of the concrete isdrastically lowered. Air contents from 30 to 80% are not uncom-mon [9]. In this way, a wide range of densities can be achievedvarying between 500 and 1600 kg/m3 [10]. Cellular concrete iscomposed of cementitious mortar surrounding disconnected ran-dom air bubbles. The air bubbles are a result of gas formed withinthe mortar or foam introduced into the mortar mixture [11]. Basedon the method of curing, cellular concrete can be classified as non-autoclaved or autoclaved [12]. When the binder consists of otherthan just portland cement, autoclave curing usually is employed[9]. Within the context of the present paper, only cellular concreteoriginating from an autoclaved process (ACC) is considered, focus-ing on the precast masonry products namely the cellular concreteblocks. As presented in Table 2, the investigated materials aremade out of very fine quartz, ground sand [13–16], lime, cementsometimes combined with fly ashes [10] and as air-entrainingagent, aluminium powder [13] in combination with a foamingagent. Recycled materials coming from rejected autoclaved andnon-autoclaved cellular concrete are already reused in the produc-tion process itself (Table 2). The aluminium powder will react withthe lime or alkaline substances which will bring hydrogen into thecementitious mortar [13], [16]. The foaming agent is used to attractthe cement particles into the aerosol foam network by which hy-drated portland cement paste is formed around each entrappedair bubble [11]. XRD powder studies have already shown that themain reaction products belong to the tobermorite group of calciumsilicate hydrates (C ± S ± H) [14–17] using the calcium from thelime and cement and the silicates from the quartz, the fine sandand cement (Table 2). The reaction products are a mixture of crys-talline, semi-crystalline and near amorphous materials with vary-ing degree of crystallinity [14,15,17]. The autoclaved cellularconcrete has many applications in building engineering, mainlyin the housing, industrial and public utility building [18].
A strategy was chosen to make the simulations and tests withthe cellular concrete ARM as realistic as possible, by using threemodern reference clinker factories. Simulations were carried outand experimental clinkers were produced to investigate the influ-ence of the grindability of the cellular concrete ARM on the miner-alogical composition of the final clinker.
2. Materials and methods
2.1. Classic raw materials (CRM)
As representative CRM, materials are selected that are used at a daily base inthree reference clinker factories. These factories are CBR Antoing (CRM/Ant) andCBR Lixhe (CRM/Lxh) in Belgium and ENCI Maastricht (CRM/Maa) in the Nether-
Table 1Average chemical analysis of the limestones and SiO2-sources of CBR Lixhe and ENCIMaastricht.
lands, all belonging to the Heidelbergcement Benelux group. CBR Antoing usestwo kinds of limestones, rich (CRM/Ant/RL) and poor (CRM/Ant/PL), CBR Lixhe usestufa (CRM/Lxh/Tu) and loam (CRM/Lxh/Lo) and ENCI Maastricht a typical marl(CRM/Lxh/Ma) and sabulous clay (CRM/Lxh/SC). All of the 3 factories use fly ash(CRM/Ant,Lxh,Maa/FA) as Al2O3 source and an artificially produced Fe2O3 source(CRM/Ant,Lxh,Maa/IC). These CRM were already described in detail [2]. The chem-ical analyses of the CRM which directly influence the current investigation namelyCRM/Lxh/Tu, CRM/Lxh/Lo, CRM/Maa/Ma and CRM/Maa/SC are presented in Table 1.
2.2. Alternative raw material (ARM): cellular concrete
The cellular concrete materials used within this study were selected from twosources, polluted recycled cellular concrete (ARM/CCP) and production waste ofclean cellular concrete (ARM/CCC). The ARM/CCP came from a demolition plantKOK in the Netherlands. Six samples spread over a period of two months weredelivered and chemically analysed (Table 3). The ARM/CCC samples were recoveredfrom the factory of Xella/Ytong Burcht in Belgium during twenty-eight weeks. Eachweek, samples were taken out of the crushed material flow of rejected cellular con-crete blocks which were re-entered in the cellular concrete production of Xella Bur-cht as described in Table 2. These rejected cellular concrete blocks were withheldbecause of their unsuitable dimensions. These samples were afterwards 2 by 2homogenised which delivered a total of fourteen samples that were further ana-lysed and investigated (Table 4). Because the process is quite sensible to the small-est changes, it was foreseen that the chemical variation was quite small. Changescould influence the rising, dimensioning and curing of the cellular concrete blocks.The ARM/CCP have some inorganic and organic contaminants which will be furtherexplained in Section 4.2. It was decided to investigate the influence and the restric-tions of cellular concrete on Portland clinker production with the ARM/CCC materi-als to avoid too many influencing variables that could make an objective evaluationimpossible. The ARM/CCP will give, based on their chemical analysis, additionalinformation about the expected chemical variation and the possible contaminationsif real recycled cellular concrete should be used as ARM for Portland clinkerproduction.
2.3. Testing of raw materials, Cold Clinker Meals (CCM) and clinker properties
To prepare the different CCM compositions, raw materials were dosed toachieve 500 g of CCM in line with the clinker feed calculation described in Sec-tion 3.3. All CCM materials were ground for 5, 10, 20, 30 min at 300 rpm in a labo-ratory ball mill to obtain different degrees of fineness. These were determined bySympatec laser diffraction. The CCM materials ground for 10 min were, after a gran-ulation phase, sintered in an electric high temperature static kiln (CarboliteBLF1800) as described in [2]. After evaluation of the mineralogical composition itwas decided to sinter the alternative CCM with a grinding time of 20 and 30 minto investigate the influence on the clinker mineralogy.
The classic raw materials loam and sabulous clay used as SiO2 source in the fac-tories of CBR Lixhe and ENCI Maastricht and the alternative raw material ARM/CCC/S3 were ground in succession for 1–10 min with an interval of 1 min in the Sieb-technic disc mill to compare their grindability. The fineness was also determinedby Sympatec laser diffraction.
XRF analyses were determined with a Philips PW2404, total C and S contentwere analysed with a Leco CS230, TGA/DTA analyses were performed using a Net-zsch STA 449F3 and finally XRD analyses refined by Rietveld method by a Bruker D8ADVANCE. The X-ray diffraction pattern was measured using a Bruker D8 ADVANCEdiffractometer. An X-ray diffraction analysis based on Topas (DIFFRAC.SUITE) pro-file and structure analysis software was used to quantify the different clinkerphases. The data was collected with a vertical Theta–Theta goniometer with�110� < 2h < 168� goniometer control using a step size of 0.0001�. Background, zeroshift, scaling factor, cell parameters and shape parameter were refined. The usedcrystal structure models are listed in Table 5.
Table 3Chemical composition of the recycled cellular concrete samples and the average of these samples.
ARM (wt%) ARM/CCP/S1 ARM/CCP/S2 ARM/CCP/S3 ARM/CCP/S4 ARM/CCP/S5 ARM/CCP/S6 Average all
Table 6Chemical and mineralogical limitations on the final clinker.
Clinker (wt%) Antoing Lixhe Maastricht
Cl x < 0.08 x < 0.08 x < 0.08SO3 x < 1.4 x < 1.2 x < 1.1Na2Oeq x < 1.2 x < 1.2 x < 1.2MgO x < 4.0 x < 4.0 x < 4.0MgO/Fe2O3 x < 1.40 x < 1.40 x < 1.40DoS-level 80 < x < 120 80 < x < 120 80 < x < 120
J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733 727
3. Theory/calculation
3.1. Chemical and mineralogical limitations of each reference clinkerand clinker kiln
To prevent undesirable effects on both the clinker productionprocess [19] as well as on the clinker quality and to remain withinthe framework of the cement standards [20,21], chemical limits forSO3, Cl, alkalis specific for each clinker factory were defined andused to evaluate the feasibility of applying the ARM/CCC in clinkerproduction (Table 6). With respect to the mineralogy of the finalclinker, limits are also defined for the following three parameters:Lime Saturation Factor (LSF), C3A and the liquid phase (LiqSimple)[19]. All these limitations were already described and explained indetail in [2]. Also two limits on MgO (Table 6) were defined as de-
scribed in [8], although not critical within this current study(Table 4).
3.2. Influence of grinding fineness on clinker production and cementhydration
Fineness and particle size distributions affect the burnability ofthe Cold Clinker Meal (CCM). As the fineness of the CCM goesdown, a surface area is created that will facilitate the sinter processand will lower the sintering temperature. This relationship isextensively described in literature [22–24]. The raw meal finenesswill have its influence particularly on the formation of the liquidphase and the growth rates of silicate crystals [22]. Especially,the fineness of quartz is found to have a very strong effect on theCCM burnability [25,26]. In fact the maximum permissible particlesize of quartz, feldspars and calcite is theoretically recommendedto be 45 lm, 63 lm and 125 lm, respectively [24] although thesevalues are not used in clinker factories as steering parameters. Tar-get finenesses of the CCM are described to be maximum 12 wt%residue on a 90 lm sieve and 2.6 wt% on a 211 lm [25]. Taking intoaccount that quartz is the most difficult to grind compared to feld-spar and calcite [8] (Mohs scale), it is quite uncertain that a particlesize below 45 lm will be obtained if the target for CCM fineness is12 wt% residue at 90 lm. In modern clinker factories such as thethree reference clinker factories, even coarser CCM finenesses withregard to the residue on 90 lm sieve, are pursued (Table 9) as alsostated by [22]. The relationship between the raw material fineness-es and the burnability of the CCM expressed as the free CaO level(wt%) of the final clinker was investigated and described mathe-matically by Fundal and Christensen [24,27,28]. Their mathemati-cal expressions are called the chemico granulometric approach andare based on chemical and physical analyses of 30 industrial rawmixtures and some steric assumptions and approximations. Amathematical expression (1) is used to describe the relationship
Table 7Compositions of the different clinker meals made to be fed to the static kiln.
CRM + ARM Quantity (wt%) CRM + ARM Quantity (wt%) CRM + ARM Quantity (wt%)
Table 9Target values (wt%) for particle size distribution of the reference CCM.
CCM (lm) Antoing Lixhe Maastricht
63 <27.0 – –90 <19.0 <16.0 <20.0
200 <3.5 <1.0 <3.5
59
60
61
62
63
CaO (wt%)
SiO
2 (w
t%)
ARM/CCC/S3
ARM/CCC/S7
ARM/CCC/S1
30 31 32 33 34
Fig. 1. CaO (wt%) in function of SiO2 (wt%) without Loss of Ignition (950 �C) of cleancellular concrete. The labels mark the selected clean cellular concrete materials.
728 J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733
between the individual raw material finenesses and theburnability.
This explains, although not used in industrial practice, quite ni-cely the importance of the quartz fineness on the CCM burnability.
3.3. Controlling parameters for clinker feed calculation
A simulation program based on linear equations, was used tocalculate Cold Clinker Meals (CCM) for each factory (CCM/Ant,Lxh,Maa) out of the CRM, in the case of the reference CCM as well aspartly out of the ARM, in the case of the alternative CCM. Thesetype of calculations were already explained in detail in [2,8].Important to notice is that a stoichiometric balance is imposedby the simulation program between SO3 and alkali which is
expressed as the Degree of Sulfatisation (DoS) value, calculatedby Eq. (8) using the chemical analysis of the final clinker.
DoS ¼ 77:41 � SO3=ðNa2Oþ K2O � 0:658Þ ð8Þ
DoS levels between 80 and 120 wt% are recommended [2] andcurrently used as process parameter in the three reference clinkerfactories. The chemical and mineralogical limitations described inSection 3.1, are also incorporated in the simulation program.
4. Results and discussion
4.1. Clinker feed calculations and preparations
The CCM were calculated with a simulation program (section3.3) in line with the chemical and mineralogical requirementslisted in Table 6. The compositions of these CCM after simulationare presented in Table 7. The reference CCM have the same compo-sition as used today in the three clinker factories. The alternativeCCM were calculated to maximise the use of the cellular concreteARM. Because of the high SiO2 (wt%) in the cellular concrete mate-rials, it was expected that they could function as SiO2-source inCCM replacing loam or sabulous clay as classic raw materials(Table 1).
From the three selected clean cellular concrete samples, theARM closest to the average SiO2 (wt%) (Fig. 1) or ARM/CCC/S3was used in the simulation program and the corresponding alter-native Cold Clinker Meal (CCM) preparation. By maximisation ofthese materials in the different Cold Clinker Meals (CCM), the loamand sabulous clay dosages (wt%) were completely replaced in the
J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733 729
case of CBR Lixhe and ENCI Maastricht (Table 7). The only limitingfactor was the SiO2 (wt%) of the CCM itself. The dosage of ARM/CCC/S3 in the alternative CCM of CBR Lixhe (CCM/Lxh/CC) wascomparable with the dosage of CRM/Lxh/Lo in the reference CCM(CCM/Lxh/Ref) and also the CaO-source CRM/Lxh/Tu was margin-ally lowered. This is in contradiction with the alternative CCM ofENCI Maastricht (CCM/Maa/CC) where the dosage of ARM/CCC/S3almost quadrupled compared to CRM/Maa/SC dosage in the refer-ence CCM (CCM/Maa/Ref) and the CaO-source CRM/Maa/Ma wassignificantly lowered. Maximisation of ARM/CCC/S3 in the CCMof CBR Antoing showed little or no possibilities to use cellular con-crete. This is due to the fact that CBR Antoing does not use a realSiO2-source as the necessary SiO2 is delivered by the two limestonesources, especially the poor limestone (CRM/Ant/CP). Decreasingthe CRM/Ant/CP dosage would nevertheless result in a too low le-vel of CaO (wt%). Therefore no CCM for CBR Antoing was made be-cause it would differ marginally from the reference. The CCM wereprepared and sintered as described in Section 2.3. The physical andchemical analysis of these CCM as well as the chemical and miner-alogical analysis of the final clinkers will be discussed in the fol-lowing paragraphs.
4.2. Chemical and TGA analysis
As can be noticed out of Table 4, the chemical analyses of theclean cellular concrete materials (ARM/CCC) were significantly dif-ferent from those originating from polluted cellular concrete(ARM/CCP) as presented in Table 3. Even when comparing Fig. 1with Fig. 2, where both cellular material sources are expressedwithout L.O.I., being the way the ARM will be incorporated in thefinal clinker, the difference is quite clear. Xella Burcht uses whitePortland cement which is quite unique in the production processof cellular concrete. The cellular concrete blocks demolished inthe Netherlands by Kok have probably a composition based on greyPortland cement or even high blast furnace slag cement which isdemonstrated by the higher level of Fe2O3 for grey Portland cementor the lower level of CaO (Figs. 1 and 2) in the case of blast furnaceslag cement compared to these of the clean cellular concrete ofXella Burcht. Furthermore, it is quite noticeable that the ARM/CCP of Kok have on average, higher levels of L.O.I., SO3 and C thanthe ARM/CCC of Xella Burcht indicating that there are some inor-ganic and organic contaminants of another origin than the cellularconcrete itself. Mainly the high levels of SO3, probably originatingfrom inorganic contaminants as plaster which could cover therecycled cellular concrete blocks, will be a serious bottleneck whenusing real recycled or polluted cellular concrete (ARM/CCP) in
53545556575859606162636465666768697071
22 23 24 25 26 27 28CaO (wt%)
SiO
2 (w
t%)
Fig. 2. CaO (wt%) in function of SiO2 (wt%) without Loss of Ignition (950 �C) ofpolluted cellular concrete.
Portland clinker production as already explained in [2]. The organiccontaminants as adhesives to glue cellular blocks together whichare demonstrated by the increased C (wt%) could be critical whencellular concrete ARM are dosed in significant amounts [2]. Thechemical variation of the cellular concrete materials can be visual-ised by comparing Figs. 1 and 2. It is clear that the ARM/CCC arechemically more stable than the ARM/CCP which is quite logicdue to the fact that the ARM/CCC is generated out of productionwaste and ARM/CCP comes from real recycled waste. When gener-ated out of recycling, cellular concrete could also originate fromdifferent production factories with other raw materials. The cellu-lar concrete could also be contaminated by other constructionmaterials when been used in construction projects. The chemicalinstability of ARM/CCP will require a homogenisation phase beforeit could be used in Portland clinker production. This will generatean extra energy cost. When comparing the ARM/CCC materialswith CRM/Lxh/Lo and CRM/Maa/SC being the classic SiO2-sourcesof CBR Lixhe and ENCI Maastricht, it is clear out of Tables 1 and4 that the ARM/CCC are significantly lower in SiO2 (wt%) and high-er in CaO (wt%) taking in account the L.O.I. (wt%) of the CRM. Thisexplains the different alternative CCM compositions mentioned inSection 4.1.
TGA-analyses of the selected ARM/CCC (Fig. 4), shows high lev-els of physically as well as chemically bound H2O. The physicallybound H2O, that is released between 25 and 100 �C, originates fromthe porous structure of the cellular concrete itself generated by thehigh concentration of air voids. The water absorption of air-curedcellular concrete materials is visualised in Fig. 3. When recycledin practice, these air voids will be filled partly with unreactedand external absorption H2O which will need energy to evaporate.The mass loss of approximately 10 wt% between 100 and 500 �Cout of the TGA-analyses of the ARM/CCC (Fig. 4), comes from thethermal decomposition of the calcium silicate hydrates [12,15–17]. Furthermore a small but distinct decarbonation peak between650 and 850 �C of approximately 3 wt% was measured. Both CRM/Lxh/Lo as CRM/Maa/SC used as classic SiO2-source in CBR Lixhe andENCI Maastricht, show small but quantifiable (between 1% and 4%)decarbonation mass losses between 650 and 850 �C (Fig. 5). The bigdifference with cellular concrete materials is that the chemicallybound H2O is only marginally present for 2–3 wt% in both CRM/Lxh/Lo as CRM/Maa/SC.
The use of cellular concrete as ARM can have an influence on thedecarbonation (wt%) of the CCM and hence CO2 emissions of theclinker kilns when it influences the CaCO3 source, which is tufa
0
10
20
30
40
50
impregnatedAsphalt treated
Wat
er A
bsor
ptio
n (w
t%)
Non AutoclavedAutoclaved
Plain Slurry Coated Sulfur
Fig. 3. Water absorption characteristics of lightweight concrete [13].
Fig. 5. TGA/DTA analysis of the classic raw materials (ARM/Lxh/Lo; ARM/Lxh/SC).
730 J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733
for CBR Lixhe and marl for ENCI Maastricht. In the case of ENCIMaastricht, the marl dosage (wt%) is lowered by the use of ARM/CCC/S3 in the alternative CCM (Table 7). Out of Table 8, it is clearthat the decrease of marl results in a slight decrease in decarbon-ation. An additional mass loss of approximately 2 wt% can be mea-sured by the dehydration of the calcium silica hydrates aspresented in Table 8 which lowers by increasing grinding time.This means that the calcium silicate hydrates are affected by grind-ing and will liberate their components by grinding. The longer thealternative CCM are ground, the lower the amount of chemicallybound H2O (wt%) and therefore the lower the calcium silicate hy-drates content (wt%) will be. The endothermal dehydration of thecellular concrete ARM will energetically be disfavoured comparedto the use of the classic SiO2-sources (CRM) used up to day in thefactories of CBR Lixhe and ENCI Maastricht. Together with the highlevels of organic and inorganic contaminants measured in theARM/CCP, cellular concrete will have some restrictions if been used
as alternative raw material in Portland clinker production. Theserestrictions will also be valid for most recycled concrete materialsif used as ARM for Portland clinker production.
4.3. XRD analysis and the mineralogical influence of Cold Clinker Mealfineness
The XRD analyses coupled with Rietveld refinement presentedin Table 11, shows different total mineralogical weight percentagesof each mineral out of its different mineralogical structures thanthose calculated by simple Bogue equations based on the chemicalanalysis presented in Table 10 of the final reference and alternativeclinkers. Although all the CCM were designed to have DoS-factorsbetween 80 and 120 (Table 6), many of the final clinkers did notachieve this goal. The reason of this unbalance between alkaliand SO3 (wt%) is due to the different volatility of the Cl, SO3 andalkali in the static lab furnace compared to a real clinker kiln [2].
Table 10Chemical analysis and Bogue calculations of the final clinkers of CBR Lixhe and ENCI Maastricht made in a static kiln.
Clinker (wt%) Cl/Lxh/Ref 10 min Cl/Lxh/CC 10 min Cl/Maa/Ref 10 min Cl/Maa/CC 10 min
J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733 731
This could generate a small difference between the simple Boguecalculation and the XRD analyses. Nevertheless, evaluation be-tween XRD analyses of the clinkers made out of the same alterna-tive CCM with different grinding times could still objectively bedone by comparing them to the corresponding Bogue equation.At first sight, as presented in Figs. 6–8 for the residue on 90 lm,63 lm and 45 lm sieves, the grindability of CCM/Lxh/CC andCCM/Maa/CC is clearly worse than that of CCM/Lxh/Ref and CCM/Maa/Ref. The residue on the 200 lm sieve was unnecessary topresent because 0 wt% was retained after a grinding time of only5 min. Based on Table 9, sufficient fineness should be attaint after10 min grinding for all CCM. However, this is not supported by the
Fig. 6. Residue on the sieve of 90 lm for all CCM after 5, 10, 15, 20 and 30 mingrinding time.
measured mineralogy of the alternative clinkers generated out ofCCM/Lxh/CC and CCM/Maa/CC. Out of Table 11, it is clear thatthe burnability of these two alternative CCM after 10 min grinding,is quite insufficient demonstrated by the free CaO (wt%) of2.08 wt% for Cl/Lxh/CC and 1.53 wt% for Cl/Maa/CC compared to0.23 wt% and 0.35 wt% for the reference clinkers. This results intoo low alite (wt%) for the alternative clinkers. On the other hand,if the residue on the 90 lm sieve has to be less than 12 wt% as sta-ted by Ghosh [25], both alternative CCM do not achieve this goal. Ifthe grinding time for the alternative CCM is increased to 20 min,the burnability becomes acceptable with free CaO (wt%) of0.96 wt% for Cl/Lxh/CC and 0.76 wt% for Cl/Maa/CC, which is
Fig. 8. Residue on the sieve of 45 lm for all CCM after 5, 10, 15, 20 and 30 mingrinding time.
0
10
20
30
40
50
60
70
80
90
100
1 min 2 min 3 min 4 min 5 min 6 min 7 min 8 min 9 min 10 min
[wt%
]
CRM/Lxh/Lo CRM/Maa/SC ARM/CCC/S3
Fig. 10. Residue on the sieve of 45 lm for the classic SiO2-sources and cellularconcrete material after 1–10 min grinding time.
732 J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733
however still higher than for the reference clinkers. This was notimproved after 30 min of grinding time although residues on the90 lm sieve (Fig. 6) and on the 63 lm sieve (Fig. 7) were furtherlowered. The reason for this deviation can be explained by compar-ing the residue on the 90 lm sieve and the 45 lm sieve of the twoindividual classic raw materials (CRM) that are used as SiO2-source, CRM/Lxh/Lo (CBR Lixhe) and CRM/Maa/SC (ENCI Maas-tricht) and the alternative raw material (ARM) or cellular concretematerial ARM/CCC/S3 (Figs. 9 and 10). Both figures have a compa-rable trend. As explained in Section 4.1 and presented in Table 7,both CRM are replaced by the ARM in the alternative CCM. As dem-onstrated in Section 3.2, the fineness of the SiO2-source is directlyrelated to the burnability of the CCM. CRM/Lxh/Lo is quite easy togrind compared to ARM/CCC/S3 and CRM/Maa/SC which explainsthe better grindability of CCM/Lxh/Ref compared to CCM/Lxh/CC.The grindability of CRM/Maa/SC is worse than this of ARM/CCC/S3 but on the other hand delivers more SiO2 to the CCM. This re-sults in the fact that ARM/CCC/S3 has to be dosed almost four timesas much in CCM/Maa/CC than CRM/Maa/SC in CCM/Maa/Ref.Therefore the poor grindability of CRM/Maa/SC does not result ina reduced grindability of the CCM/Maa/Ref compared to CCM/Maa/CC even at the contrary due to the lower dosage (wt%). Thepoor grindability of ARM/CCC/S3 in function of its SiO2 (wt%) canbe explained out of Table 2. The ground sand and the very finequartz form together with the cement and the lime after autoclav-ing, tobermorite compounds of calcium silicate hydrates
0
10
20
30
40
50
60
70
80
90
100
[wt%
]
CRM/Lxh/Lo CRM/Maa/SC ARM/CCC/S3
1 min 2 min 3 min 4 min 5 min 6 min 7 min 8 min 9 min 10 min
Fig. 9. Residue on the sieve of 90 lm for the classic SiO2-sources and cellularconcrete material after 1–10 min grinding time.
(C ± S ± H). Tobermorite compounds are quite easy to grind (Mohs2,5). On the other hand, quartz particles are more difficult to grind(Mohs 7). The quartz particles coming from the fine quartz have avery high fineness of 6500 cm2/g, more than sufficient for ideal sin-tering conditions for Portland clinker production. The ground sandhas a much lower fineness with a 15–25 wt% residue on 90 lm.Unground residues (wt%) bigger than 90 lm (Fig. 9) will be deliv-ered by the tobermorite compounds, coming from the autoclavingprocess and the re-entered autoclaved and nonautoclaved recov-ered cellular concrete materials (Table 2). Also noncombinedquartz sand particles will deliver partly residues bigger than90 lm. During grinding, the calcium silicate hydrates will increas-ingly disconnect their components by increasing grinding time asexplained in Section 4.2, liberating part of the fine and coarsequartz particles. The coarse quartz particles coming from theground sand (Table 2), have to be further ground to attain the nec-essary fineness. It can be concluded, that the grindability of ARM/CCC/S3 will be a combination of the grindability of the tobermoritecompounds and the grindability of the coarser quartz particles.Although both CRM as also ARM/CCC/S3 consist out of quartz par-ticles, it is the presence of the tobermorite compounds in the ARM/CCC/S3 that generates the poor grindabilty of the ARM/CCC/S3 infunction of the delivered SiO2 (wt%). As explained by Fundal andChristensen [24,27,28], quartz particles bigger than 45 lm willhave a major influence on the burnability of the CCM which ex-plains why the alternative CCM even after 20 and 30 min grindingtime have an improved but relatively high free CaO (wt%). There-fore it can be stated that monitoring CCM fineness by residueson 90 lm and 211 lm is not sufficient to guarantee good burnabil-ity when using ARM with poor grindability in function of their SiO2
(wt%) which replaces CRM as SiO2-source in Portland clinker pro-duction. In the case of other ARM as for example fibrecementmaterials [2] where also calcium silicate hydrates are present,the reduced grindability was not seen because the quartz particlesoriginated from cementitious materials which have already suffi-cient fineness (4000 cm2/g). Nevertheless, the demonstrated bot-tleneck with the quartz grindability will be valid for all ARMcomposed out of high levels of coarse quartz sand as most recycledconcrete materials.
5. Conclusions
As could be noted in the different paragraphs, cellular concretematerials could be used as an Alternative Raw Material (ARM) forPortland clinker production. It was proven that they could replaceinert Classic Raw Materials (CRM) as sabulous clay or loam which
J. Schoon et al. / Construction and Building Materials 48 (2013) 725–733 733
are used as SiO2-source. On the other hand, the replacement of theCRM by the cellular concrete ARM will not have major beneficialadvantages. The physical and chemical bound H2O, which is pres-ent at a larger extent in the ARM than in the CRM, will consumemore energy during evaporation. Also the presence of the quartzsand and their incorporation in calcium silicate hydrates, couldact as a serious restriction by the increased grinding efforts needed,compared to classic raw materials. Although not insurmountable,the energy necessary to grind this fraction in function of the deliv-ered SiO2, limits the use of recycled cellular concrete as ARM in anenergetic and ecological way. Also possible contaminants as plas-ter, gluing mortars or organic glues and the quite big chemical vari-ety of real recycled cellular concrete renders its use as ARM,although possible, unpractical. A positive effect on the ecologicalimpact of Portland clinker production is that a primary naturalmaterial could be replaced by a secondary recycled material whichis in line with the Cement Sustainability Initiative. Although thepossible use of cellular concrete in Portland clinker production isnot a success story overall, this feasibility study reveals neverthe-less some of the restrictions which could be generated when usingrecycled concrete materials as alternative raw material for Port-land clinker production.
Acknowledgements
The authors wish to thank Zjelko Rudic of Xella Burcht and thecement research lab of ENCI Maastricht for their support, Els Bru-neel for her aid during the practical execution of the TGA tests,Jo Lejeune for his aid during the XRD tests and Tanguy Ewbankfor the support during laser diffraction measurements.
References
[1] Damtoft JS, Lukasik J, Herfort D, Sorrentino D, Gartner EM. Sustainabledevelopment and climate change initiatives. Cem Concr Res 2008;38:115–27.
[2] Schoon J, Van der Heyden L, Eloy P, Gaigneux EM, De Buysser K, Van DriesscheI, et al. Waste fibrecement: an interesting alternative raw material for asustainable Portland clinker production. Constr Build Mater 2012;36:391–403.
[3] Puertas F, García-Díaz I, Barba A, Gazulla MF, Palacios M, Gómez MP, et al.Ceramic wastes as alternative raw materials for Portland cement clinkerproduction. Cement Concr Compos 2008;30:798–805.
[4] Pardo N, Moya JA, Mercier A. Prospective on the energy efficiency and CO2
emissions in the EU cement industry. Energy 2011;36:3244–54.
[5] Mikulcic H, Vujanovic M, Duic N. Reducing the CO2 emissions in Croatia’scement industry – the precalciner model. In: 6th Dubrovnik conference onsustainable development of energy, water and environment systems; 2011.
[6] Baier H, Menzel K. Proven experiences with alternative fuels in cement kilns.In: 13th International congress on the chemistry of cement; 2011. p. 1–7.
[7] Hojamberdiev M, Muhamedbaeva Z, Madhusoodana CD. Use of natural andthermally activated porphyrite in cement production. Constr Build Mater2009;23:2757–62.
[8] Schoon J, Vergari A, De Buysser K, Van Driessche I, De Belie Nele. Finesextracted from porphyry and dolomitic limestone aggregates production: MgOas fluxing agent for a sustainable Portland clinker production. Constr BuildMater 2013;43:511–22.
[9] Hoff George C. Porosity-strength considerations for cellular. Cem Concr Res1972;2:91–100.
[11] Tikalskya PJ, William MacDonald JP. A method for assessment of the freeze–thaw resistance of preformed foam cellular concrete. Cem Concr Res2004;34:889–93.
[12] Narayanan N, Ramamurthy K. Structure and properties of aerated concrete: areview. Cement Concr Compos 2000;22:321–9.
[13] Mirza WH, AI-Noury SI. Utilisation of Saudi sands for aerated concreteproduction. Int J Cement Compos Lightweight Concr 1986;8(2).
[14] Isu N, Ishida H, Mitsuda T. Infuence of quartz particle size on the chemical andmechanical properties of autoclaved aerated concrete (I) tobermoriteformation. Cem Concr Res 1995;25:243–8.
[15] Isu N, Ishida H, Mitsuda T. Infuence of quartz particle size on the chemical andmechanical properties of autoclaved aerated concrete (II) fracture toughness,strength and micropore. Cem Concr Res 1995;25:249–54.
[16] Alexanderson J. Relations between structure and mechanical properties ofautoclaved aerated concrete. Cem Concr Res 1979;9:507–14.
[17] Mitsuda T, Chan CF. Anomalous tobermorite in autoclaved aerated concrete.Cem Concr Res 1977;7:191–4.
[18] Beben D, Manko Z. Influence of selected hydrophobic agents on someproperties of autoclaving cellular concrete (ACC). Constr Build Mater2011;25:282–7.
[19] Taylor HFW. Cement chemistry, vol. xviii. 2nd ed. London: T. Telford; 1997.[20] EN 197-1. Cement – part 1: composition, specifications and conformity criteria
for common cements; 2000.[21] EN 197-2. Cement – part 2: conformity evaluation; 2000.[22] Kakali G, Tsivilis S. The effect of intergrinding and separate grinding of cement
rwa mix on the burning process. Cem Concr Res 1993;23:651–62.[23] Tsamatsoulis D. Modeling of raw materials blending in raw meal grinding
systems. Latest trends on systems, vol. I. ISSN: 1792-4235.[24] Christensen NH. Burnability of cement raw mixes at 1400 �C, the effect of the
fineness. Cem Concr Res 1979;9:285–94.[25] Ghosh SN. Progress in cement and concrete: part 1. Burnability and
clinkerization of cement raw mixes, vol. 1. Thomas Telford; 1991.[26] Bogue RH. The chemistry of cement. New York: Reinhold Publishing Corp;
1960.[27] Fundal E. World cement technologie 1979;10:195–6 (199–200, 203–204).[28] Christensen NH. Burnability of cement raw mixes at 1400 �C, the effect of the
