Construction and Building Materials 43 (2013) 511–522

Contents lists available at SciVerse ScienceDi rect

Construc tion and Buildi ng Materi als

journal homepage: www.elsevier .com/locate /conbui ldmat

Fines extracted from porphyry and dolomitic limestone aggregates

production: MgO as fluxing agent for a sustainable Portland clinker production

0950-0618/$ - see front matter � 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2013.02.046

Abbreviations: Ant, Antoing; Ag, aggregates; ARM, alternative raw material; ARM/Pp, porphyry aggregates; ARM/PpF, porphyry fines; ARM/DL, dolomitic limestone; ARM/DLS, dolomitic limestone sludge; CCM, cold clinker meal; Cl, clinker; CRM, classic raw material; Decarb E, decarbonation energy; DoS, degree of sulfatisation; DLSl, dolomitic limestone sludge; FA, fly ash (Al2O3-source); HCM, hot clinker meal; IC, iron carrier (Fe2O3-source); LSF, lime saturation factor; LiqSimple, liquid simple; LOI, loss on ignition; Lo, loam (SiO2-source); Lxh, Lixhe; Maa, Maastricht; Ma, marl (specific type of limestone); PpD, porphyry dust; PL, poor limestone; Ref, reference; RL, rich limestone; SC, sabulous clay (SiO2-source); SR, saturation rate; Tu, tuffeau (specific type of limestone).⇑ Corresponding author. Tel.: +32 9 264 55 22; fax: +32 9 264 58 45.

E-mail address: nele.debelie@UGent.be (N. De Belie).

Joris Schoon a,b, Anne Vergari a, 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

h i g h l i g h t s

� Porphyry fines could replace fly ash in a Portland clinker meal. � MgO could have a positive mineralogical influence on Portland clinker produc tion. � MgO in combination with alkali and SO 3 has fluxing activities. � The fluxing effect promotes the alite formation. � A decrease of the liquid formation temperature is noticeable by the fluxing effect. � Finetuning of MgO (wt%) up to 2.0–2.5 wt% in Portland clinker could be recommended.

a r t i c l e i n f o

Article history: Received 10 November 2012 Received in revised form 31 January 2013 Accepted 26 February 2013 Available online 2 April 2013

Keywords:ClinkerPorphyryDolomitic limestone FinesSludgeAl2O3

MgOCementFly ash Alite

a b s t r a c t

This paper aims to examine the use of fines and sludge generated out of the production of porphyry and dolomitic limestone as an alternative raw material for Portland clinker kilns with enumeration of possible limitations. The possibility to generate a raw material with a stable compositional variation was investi- gated and simulation s were carried out to maximise their use in clinker kilns. Furthermore, experimental clinkers were produced with dosages that were esteemed as realistic by the numerical simulations. The final clinkers were fully analysed and evaluated on possible mineralogical influences.

� 2013 Elsevier Ltd. All rights reserved.

1. Introductio n

The Cement Sustainability Initiative [1] defined some indicators as energy reduction s as well as an increasing alternative raw mate- rial use as key issues in the sustainable developmen t of the cement industry [2]. Significant research on the use of industrial waste and by-produ cts as Alternative Raw Materials (ARMs) to partially re- place conventional or Classic Raw Materials (CRMs) has already been performed [3–5]. The recuperatio n of fines and sludge out

512 J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522

of aggregates production and also their use as replacemen ts of CRM for the production of Portland clinker could minimise the ef- fects of quarrying in both production lines as well as the need to grind them as raw materials. Furthermore a positive influence on the Life Cycle Assessment (LCA) and carbon footprint could be measured for concrete and asphalt production when fine particles (wt%) are reduced in the aggregates that are used. Sludge or finescoming from the production of pure limestone aggregat es could easily be recovered in clinker kilns by their chemical similarity with the classic limestone raw materials. A less obvious choice is sludge or fines coming from the production of porphyry or dolomi- tised limestone aggregat es. The studied ARM could replace CRM which have higher energy consumption by the quarrying process or which could be valorised for a nobler use in line with their intrinsic hydraulic reactivity. Furthermore, it is shown that these less obvious Alternative Raw Materials (ARMs) for Portland clinker production have some additional advantag es because of their high- er MgO content (wt%). MgO is described in literature as a possible fluxing agent which could have mineralogical influences improvin gthe burnability as well as increasing the alite formation of the finalclinker. All these properties will have a positive effect on the sus- tainability of Portland clinker production.

2. Materials and methods

2.1. Classic Raw Materials (CRMs)

As representative CRM, materials are selected that are used at a daily base in three reference clinker factories. These factories are CBR Antoing (CRM/Ant) and CBR Lixhe (CRM/Lxh) in Belgium and ENCI Maastricht (CRM/Maa) in the Nether- lands, all belonging to the Heidelberg Benelux group. They could be considered as to be representative for modern clinker factories worldwide. CBR Antoing uses two kinds of limestones (Rich (CRM/Ant/RL) and Poor (CRM/Ant/PL)), CBR Lixhe uses Tufa (CRM/Lxh/Tu) and Loam (CRM/Lxh/Lo) and ENCI Maastricht a typical Marl (CRM/Maa/Ma) and Sabulous Clay (CRM/Maa/SC). All of the three factories use Fly Ash (CRM/Ant,Lxh,Maa/FA) as Al 2O3 source and an artificially produced Fe 2O3

source (CRM/Ant,Lxh,Maa/IC). These CRM were already described in detail [3].The chemical analyses of the CRM directly influencing the current investigation are presented in Table 1.

2.2. Alternative Raw Material (ARM): porphyry and dolomitic limestone

The ARM were selected from materials coming from quarries belonging to the Heidelberg Benelux group. Porphyry materials were sampled out of the quarry of Sagrex Quenast, the dolomitic limestone materials out of the quarry of Sagrex Chanxhe.

Porphyry deposits are formed when a column of rising magma is cooled in the earth crust. It consists of large-grained crystals, such as feldspar or quartz dispersed in a fine-grained feldspathic matrix. Feldspar and the feldspathic matrix consist mineralogically out of KAlSi 3O8, NaAlSi 3O8 and CaAl 2Si2O8. Also so called hornblende is found which is an isomorphous mixture of three molecules; cal- cium-iron-magnesium silicate, aluminium–iron–magnesium silicate and iron–magnesium silicate or in general (Ca, Na)2�3 (Mg, Fe, Al)5 (Al, Si)8 O22 (OH,F)2. Interesting to mention is that the minerals feldspar and quartz have higher hardness’s measured on the Mohs scale than the mineral hornblende. Porphyry is

Table 1Average chemical analysis of the limestones of CBR Antoing, CBR Lixhe and ENCI Maastric

CRM CRM/Ant/PL CRM/Ant/RL CRM/Lxh/Tu

CaO (wt%) 42.9 50.1 51.8 SiO 2 (wt%) 15.1 6.4 4.7 Al 2O3 (wt%) 2.2 0.9 0.4 Fe 2O3 (wt%) 0.9 0.4 0.3 K2O (wt%) 0.68 0.21 0.07 Na 2O (wt%) 0.25 0.25 0.02 SO 3 (wt%) 0.90 0.57 0.09 MgO (wt%) 1.1 0.9 0.7 Cl (wt%) – – 0.011 LOI 975 �C (O2) (wt%) 35.04 40.18 42.03

used as aggregate for road building and as ballast for railroad construction due to its high compressive strength and wear resistance. Porphyry materials selected for analysis were sampled at four different places, namely two stocks of 0/20 aggre- gates, the first formed after a primary (first) crushing by a jaw and gyrator crusher (Ag/Pp020FC) and the second after a secondary crushing by a cone crusher (Ag/Pp020SC). Furthermore a stock 0/4 (Ag/Pp04Sa) sand and 0/2 (Ag/Pp02Sa) sand were sampled. By taking these samples at different places in the production process in time spread over 2011, the influence of the particle size distribution on the chemical composition can be investigated. The chemical variation of the porphyry materials is graphically presented in Fig. 1. A sample from each stock was randomly selected to be further investigated according to tests described in Section 2.3: Ag/ Pp04Sa/S5, Ag/Pp02Sa/S9, Ag/Pp020FC/S6 and Ag/Pp020SC/S9. Furthermore out of these four samples, the sample with the highest content of fines, logically the 0/2 sand (Ag/Pp02Sa/S9) was wet screened on a sieve of 63 lm to separate the finesfraction (ARM/PpF/S9) from the sand fraction bigger than 63 lm.

Known by industrial practice, this fines fraction can be dry screened with a sta- tic separator. Between 14 and 20 wt% of fines could be extracted in this way. The chemical analyses of both fractions together with the original 0/2 sand (Ag/Pp02Sa/S9) are presented in Table 2.

Dolomitic limestone is a type of limestone composed of calcite (CaCO3) and dolomite (CaMg(CO3)2). Because limestone is a sedimentary rock, the ratio be- tween dolomite and calcite can vary widely as can be noticed in Fig. 2 wherebythe content of dolomite can go up to 50 wt%. Dolomitic limestone is used as aggregate in all kinds of concrete applications. Micro-Deval tests [6] show that dolomitic limestone becomes easier to grind by increasing CaMg(CO3)2 (wt%)compared to pure limestone. The dolomitic limestone materials were also recov- ered at four different places in time spread over 2011, namely one coming from astock of 0/20 aggregates, two coming from two stocks of 0/2 sand, where one was washed and the other unwashed and finally a stock of sludge (Ag/DLSl) com- ing from the washing of the above described sands. As seen from Fig. 2, the chemical variation is quite large. Three samples of sludge, Ag/DLSl/S4b, Ag/ DLSl/S5, Ag/DLSl/S11 were selected to be analysed according to the tests de- scribed in Section 2.3. These samples were also mixed in equal proportions after which the further to be investigated ARM was formed. The chemistry of this mix (ARM/DLSl/Mix) was calculated and presented in Table 3. Porphyry neither dolo- mitic limestone is to our knowledge used up today in Portland clinker produc- tion. On the other hand, the use of natural and thermally activated porphyrite by Hojamberdiev et al. [7] as also the use of dolomitic limestone by Tsivilis et al. [8] was already investigated as raw material for Portland cement production.

2.3. Testing of raw materials, Cold Clinker Meals (CCMs), clinker and cement properties

The different CCM compositions were all prepared as described in [3]. The different raw materials were crushed in a Siebtechnic Disc mill and homoge- nised in a vessel used for the analysis of the micro-Deval abrasion resistance. All CCM were after a granulation phase, sintered in an electric high tempera- ture static kiln (Carbolite BLF1800) up to 1450 �C at a constant heating rate (10 �C/min). The reference CCM and the CCM with the dolomitic limestone ARM were also sintered at 1300 �C, 1350 �C and 1400 �C to investigate the im- pact of the high levels of MgO of the dolomitic limestone on the burnability of the clinker. The Hot Clinker Meals (HCMs) were maintained for 1 h at the scheduled temperature after which they were immediately cooled to room temperature at open air to form the final clinker. XRF on a Philips PW2404, to- tal C and S determination by a Leco CS230, TGA/DTA with a Netzsch STA 449F3 and finally XRD analysis (Bruker D8 ADVANCE) refined by Rietveld method were used according to the methods described in [3]. The concrete mixes de- scribed in Table 4 which are used to determine the influence of fines extraction on the relative water demand of the concrete mixes, were composed to have acomparable consistency S3 [33].

ht.

CRM/Maa/Ma CRM/Ant/FA CRM/Lxh/FA CRM/Maa/FA

50.8 4.8 13 5.3 7.1 53.1 46.4 50.8 0.8 20.7 18.6 23.0 0.4 7.6 7.0 7.4 0.13 1.98 1.93 2.36 0.20 0.87 0.57 0.97 0.21 0.34 0.57 1.07 0.8 1.6 1.4 1.8 – – – –40.18 16.00 7.00 4.00

15

16

57 58 59 60 61 62 63 64 65SiO2

(wt%)

Al2

O3

(wt%

)

0/20 (Jaw+Gyrat Crush) 0/20 (+Cone Crusher) 0/2 Sand 0/4 Sand Fines

Ag/Pp020FC/S6

ARM/PpS04Sa/S5

Ag/PpS02Sa/S9

Ag/Pp020SC/S9

ARM/PpF/S9

Fig. 1. SiO2 (wt%) in function of Al 2O3 (wt%) without Loss on Ignition (950 �C) of Pophyry aggregates. The arrow marks the selected Porphyry materials. (White and grey label: analysed samples; Grey label: samples used for clinker calculation and preparation.)

23456789

10111213141516171819

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50CaO

MgO

Aggregate 0/20 Unwashed Sand Washed Sand Sludge

Ag/DLSl/S5Ag/DLSl/S4b

Ag/DLSl/S11

Fig. 2. Cao (wt%) in function of MgO (wt%) without Loss of Ignition (950 �C) of dolomitic limestone aggregates.

Table 2Chemical composition of TM/QuM1/S9 and his fractions bigger and smaller than 63 lm.

ARM Porphyry 0/2 Porphyry 0/2 (>63 lm) Porphyry Fines (<63 lm)Ag/Pp02Sa ARM/PpF/S9

CaO (wt%) 3.37 3.50 3.06 SiO 2 (wt%) 61.90 62.34 57.32 Al 2O3 (wt%) 15.58 15.45 15.94 Fe 2O3 (wt%) 6.70 6.23 9.36 K2O (wt%) 2.39 2.41 2.14 Na 2O (wt%) 3.58 3.77 3.28 SO 3 (wt%) 0.07 0.10 –S (wt%) – – 0.06 MgO (wt%) 2.90 2.80 4.24 TiO 2 (wt%) 0.79 0.73 1.03 P2O5 (wt%) 0.14 0.13 0.24 Cl (wt%) – 0.01 0.01 LOI 975 �C (O2) (wt%) 2.29 2.20 2.94 Ctotal (wt%) 0.05 0.04 0.07 Stotal (wt%) 0.03 0.03 0.33

J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522 513

Table 5Chemical and mineralogical limitations on the final clinker.

Clinker Antoing Lixhe Maastricht

Cl (wt%) x < 0.08 x < 0.08 x < 0.08 SO 3 (wt%) x < 1.4 x < 1.2 x < 1.1 Na 2Oeq (wt%) x < 1.2 x < 1.2 x < 1.2 MgO (wt%) x < 4.0 x < 4.0 x < 4.0 MgO/Fe 2O3 (wt%) x < 1.40 x < 1.40 x < 1.40 DoS-level (wt%) 80 < x < 120 80 < x < 120 80 < x < 120

If [MgO] < 2 wt%

LSF_MgO (wt%) 98.24 98.19 98.20 C3A (wt%) 7.35 6.65 7.33 LiqSimple (wt%) 19.18 22.73 22.97

If [MgO] > 2 wt%, the remainder is compensated percentually (x –%MgO)

CaO (wt%) 65.92–1.51 66.56–0.78 –SiO 2 (wt%) 21.87–0.49 21.76–0.31 –Al 2O3 (wt%) 4.45–0.10 4.85–0.12 –Fe 2O3 (wt%) 2.60–0.06 3.63–0.09 –

Fig. 3. Scheme of MgO hydration with volume expansion in concrete.

Table 3Chemical composition of the three selected dolomitic limestone sludges and their mix.

ARM Ag/DLSl/S11 Ag/DLSl/S4b Ag/DLSl/S5 ARM/DLSl/Mix

CaO (wt%) 46.45 44.94 39.67 43.69 SiO 2 (wt%) 4.22 1.91 2.97 3.03 Al 2O3 (wt%) 1.89 0.63 0.85 1.12 Fe 2O3 (wt%) 1.39 0.63 1.20 1.07 K2O (wt%) 0.42 0.17 0.15 0.25 Na 2O (wt%) 0.05 0.01 0.01 0.02 SO 3 (wt%) – – – –S (wt%) 0.37 0.20 0.06 0.21 MgO (wt%) 3.50 7.68 11.30 7.49 TiO 2 (wt%) 0.09 0.03 0.05 0.06 P2O5 (wt%) 0.04 0.04 0.02 0.03 Cl (wt%) – –LOI 975 �C (O2) (wt%) 41.27 43.62 43.52 42.80 Ctotal (wt%) 10.49 12.41 11.95 11.62 Stotal (wt%) 0.42 0.12 0.06 0.20

Table 4Influence of fines on concrete composit ion and characteri stics.

Reference Ag/Pp02Sa/S9 (<63 lm included)

Ag/Pp02Sa/S9 (<63 lm not included)

CEM III/B 42.5 N LH (kg/m3) 320 320 320 Water (kg/m3) 160 160 160 River sand 0/2 (kg/m3) 705 0 0Porphyry 0/2 (kg/m3) 0 705 0Porphyry 0/2 > 63 lm (kg/m3) 0 0 705 Gravel 2/8 (kg/m3) 604 604 604 Gravel 8/16 (kg/m3) 578 578 578 Additional water (kg/m3) 0 66 40 Slump (mm) 105 110 100 Flow (mm) 420 440 440

514 J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522

3. Theory/calculati on

3.1. Chemical and mineralogical limitation s of each reference clinker and clinker kiln

To prevent undesirable effects on both the clinker production process as well as on the clinker quality and to remain within the framework of the cement standards [9,10], chemical limits for SO3, Cl and alkalis specific for each clinker factory were definedand used to evaluate the feasibility of applying the ARM, ARM/ PpF and ARM/DLSl in clinker production (Table 5). With respect to the mineralogy of the final clinker, limits are also defined for the fol- lowing three parameters: Lime Saturation Factor (LSF), C3A and the liquid phase (LiqSimple) [11]. All these limitatio ns were already de-

scribed in detail in [3] except for the part of the limitation on MgO, considerabl y present in both ARM. The average MgO (wt%) in CCM used in European clinker plants is 1.05 wt% [12]. MgO has to be monitore d because of its limitation within the cement standards [9,10] and because of the risk related to the unsoundness of con- crete [13] which will be further explained in Section 3.2. In general, MgO (wt%) is limited on Portland cement and clinker to a maximum between 4.0 and 6.0 wt% depending on the countries standardisa -tion [9,14–16] although some researchers (Gebauer, 1984) claimed that MgO up to 7.5 wt% induces no evidence of deterioration [13]. It was clear out of [13] that to achieve these high levels of MgO with- out deleterious influences, specific and not always attainable pro- cess parameters as calcining conditions, kiln temperature, residence time, cooling rate, etc. [17] are required. Therefore a lim- itation of 4.0 wt% was maintained in the simulation program used for the calculation of the different CCM. Also a limitation has to be set for the ratio between MgO and Fe 2O3 in clinker. This ratio plays an important role in the possible unsoundness of concrete. If the ratio is higher than 1.4, unsoundness is very likely [13]. Both limitatio ns were also incorporated in Table 5.

3.2. Influence of MgO on clinker production and cement hydration

MgO in clinker and cement is generally related to the soundness of concrete generally considered to be the resistance to swelling

0

1

2

3

4

0 1 2 3SO3 [wt%]

MgO

[wt%

]

M1 M3+M1 M3

Fig. 4. M1 and M3 alite formation in function of the MgO and SO 3 content wt% in Portland clinker [23].

J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522 515

resulting from an expansive chemical reaction, due to the presence of dead-burnt free CaO or MgO [18]. Within this investigatio n, MgO will especially be supplied by the dolomitic limestone sludge but will also be present in significant wt% in porphyry fines. MgO and SO 3 are the most common minor components in Portland clin- ker. MgO will especially be built in the alite phase to about max. 1.5–2.0 wt% [19,20] whereby MgO simply replaces partly the CaO. The part of the MgO which is not present in the alite phase will crystallise cubically as periclase as of 2.5 MgO (wt%) in clinker [19]. Periclase present in cement, tends to hydrate after setting un- der formation of brucite (Mg(OH)2), a condition which causes excessive expansion and possible disruption of concrete (Fig. 3)which could be measured by standardized methods as EN 196-3 [21]. Alite occurs in two monoclinic forms: M1 and M3. It is as- sumed that MgO promotes the stable growth of alite into small crystals in favour of the monoclinic form M3 [11,22,23]. The rela- tion between MgO (wt%) and SO 3 (wt%) and the formation of M1

and M3 is visualised in Fig. 4 [11] [23]. The minimum amount of MgO (wt%) necessary to stabilise M3 is estimated to be 0.6–0.8 wt% whereas in clinkers with more than 1.5 wt% of MgO, alite consists essentiall y out of M3. Although CaF 2 is the best known mineraliser [24–26], MgO has also significant fluxing activities. It could lower the temperature of the liquid formation and decrease the viscosity of the liquid phase by which the residual free lime tend to reduce [19,27,28]. Without addition of CaF 2, a combined ef- fect of R2O (R = Li, Na, K), MgO and SO 3 seems to have the most promising fluxing effect [29] which could lower the temperature of the liquid formation and significantly increases the rate of C3Sformation at 1350 �C [27]. It is important to emphasise that C3Sis not formed at temperatures below approximat ely 1200 �C [30].

3.3. Controlling parameters for clinker feed calculatio n

A simulation program based on linear equations, was used to calculate Cold Clinker Meals (CCMs) for each factory (CCM/An-t,Lxh,Maa) out of the CRM, in the case of the reference CCM as well as partly out of the ARM, in the case of the alternative CCM. The way these compositi ons were calculated as also how volatiles as SO3, Cl and alkali coming from the CRM as well as the ARM can

influence the operation of the kiln and how to deal with this, is ex- plained in detail in [3]. Additionally , the increasing levels of MgO entered by the ARM have to be mastered if higher than 2 wt%. As explained in Section 3.2, MgO will be built in the alite structure up to 2.0 wt% from which the remainder will form periclase. The LSF_MgO governs the ratio of alite and belite (1). It holds a correc- tion for the incorporati on of MgO in C3S.

LSF MgO ¼ 100 � ðCaOþ 0:75 MgO Þ � ð2:8 � SiO2 þ 1:18

� Al2O3 þ 0:65 � Fe2O3Þ ð1Þ

As stated by Taylor [11], these equation have to be altered if MgO (wt%) is bigger than 2 wt% into (2):

LSF MgO ¼ 100 � ðCaOþ 1:5Þ=ð2:8 � SiO2 þ 1:18 � Al2O3

þ 0:65 � Fe2O3Þ ð2Þ

The remainder of MgO (wt%) above 2 wt% or the part which will form periclase was percentually subtracted from the four critical metal oxides (wt%) achieved in clinker when MgO (wt%) is 2 wt% and the mineralogical limitations are in line with those described in Table 5. In this way, the ratio between these metal oxides (wt%) is maintain ed as well as the ratio between the different ma- jor constituent phases.

Formulae (3)–(6) for the percentage of Liquid (%Liq) for various combinati ons of temperature and Al 2O3/Fe2O3 are [11]:

1450 �C %Liq ¼ 3:00 � Al2O3 þ 2:25 � Fe2O3 þMgOþ K2O

þ Na2O ð3Þ

1400 �C %Liq ¼ 2:95 � Al2O3 þ 2:20 � Fe2O3 þMgOþ K2O

þ Na2O ð4Þ

1338 �C ðAl2O3=Fe2O3 P 1:38Þ %Liq

¼ 6:10 � Fe2O3 þMgOþ K2Oþ Na2O ð5Þ

1338 �C ðAl2O3=Fe2O3 6 1:38Þ %Liq

¼ �5:22 � Fe2O3 þMgOþ K2Oþ Na2O ð6Þ

This indicates why MgO and alkali works together in a com- bined fluxing effect whereby the %Liq is increased. As stated in [11], alkali sulphates decreases the viscosity of the melt, increasing the alite formatio n which explains the presence of SO 3 in the com- bined fluxing effect. When there is not sufficient SO 3, free alkali will be formed, increasing the viscosity of the melt which will low- er alite formatio n [11]. Therefore a stoichiomet ric balance is im- posed between SO 3 and alkali, expressed as the so called Degree of Sulfatisation (DoS) value calculated by Eq. (7) using the chemical analysis of the final clinker.

DoS ¼ 77:41 � SO3=ðNa2Oþ K2O � 0:658Þ ð7Þ

DoS levels between 80 and 120 wt% are recommend ed and used at present-day in the three clinker factories as process parameter. Furthermore, it was stated by Taylor [11] that the burnability of the CCM or the ease by which free lime can be reduced in the kiln to an acceptable value, decrease s with increasing LSF (1-2) or Sat- uration Rate (SR) (8).

SR ¼ SiO2=ðAl2O3 þ Fe2O3Þ ð8Þ

Increase in LSF (1–2) implies more CaO has to react, increase in SR implies less liquid at a given temperature which are both ener- getically unfavour able. Increase of MgO could have a positive influ-ence on the burnability by lowering the CaO (wt%) without an increase in SR and an increase of the liquid phase.

516 J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522

4. Results and discussion

4.1. Influence of fines extraction on aggregate application and chemistry

As mentioned in Section 2.2, the sample Ag/Pp02Sa/S9 was wet screened on a sieve of 63 lm. As can be noticed in Table 4, the extraction of fines (<63 lm) has a positive influence on the water demand of the sand when used in concrete. Compared with the water demand of a concrete based on round river sand 0/2 (ref),broken porphyry sand 0/2 (Ag/Pp02Sa/S9) increases the water de- mand with 42 wt%, porphyry sand 0/2 (Ag/Pp02Sa/S9) where finesare extracted has only a 25 wt% higher water demand. Also asphalt producers need the perfect ratio between bitumen and fines [31]whereby for energetic and technical raisons, they rather like to add than extract fines to attain the perfect ratio. Also regional stan- dardisation demands [32] to limit the wt% of the fraction smaller than 63 lm (fines) in normalised aggregates. This explains why aggregates producers extract fines by wet or dry screening if the le- vel of fines is too high by which alternative raw materials as ARM/ PpF/S9 or ARM/DLSl/Mix could be generate d.

4.2. Clinker feed calculations and preparations

The CCM were calculated with a simulation program (Sec-tion 3.3) in line with the chemical and mineralogical requirements listed in Table 5. The compositions of these CCM after simulation are presented in Table 6. The reference CCM have the same compo- sition as used today in the three clinker factories. The alternative CCM were calculated to maximise the use of the porphyry and dolomitic limestone ARM. Because of the high Al 2O3 (wt%) in porphyry materials , it was expected that they could function as Al2O3-source in CCM replacing fly ash as classic source. From the four selected samples, the ARM closest to the average SiO 2 (wt%)of all sampled porphyry materials, the porphyry sand 0/4 (Ag/Pp04Sa/S5) was used in the simulation program as well as the por- phyry fines (ARM/PpF/S9). By maximisation of these materials in the different CCM, the fly ash dosage (wt%) is lowered and some- times even replaced completely in the case of CBR Antoing and

Table 6Compositions of the different clinker meals made to be fed to the static kiln.

CRM + ARM Quantity (wt%) CRM + ARM

CCM/Ant/Ref CRM/Ant/CP 55.25 CCM/Lxh/Ref CRM/Lxh/TCRM/Ant/CR 37.50 CRM/Lxh/LCRM/Ant/FA 6.38 CRM/Lxg/FCRM/Ant/IC 0.87 CRM/Lxh/IARM 0.00 ARM Sum 100.00 Sum

CCM/Ant/Pp CRM/Ant/CP 2.81 CCM/Lxh/Pp CRM/Lxh/TCRM/Ant/CR 83.37 CRM/Lxh/LCRM/Ant/FA 0.00 CRM/Lxh/FCRM/Ant/IC 0.76 CRM/Lxh/IAg/Pp04Sa/S5 13.86 Ag/Pp04SaSum 100.00 Sum

CCM/Ant/PpD CRM/Ant/CP 23.30 CCM/Lxh/PpD CRM/Lxh/TCRM/Ant/CR 64.93 CRM/Lxh/LCRM/Ant/FA 0.09 CRM/Lxh/FCRM/Ant/IC 0.26 CRM/Lxh/IARM/PpF/S9 11.42 ARM/PpF/SSum 100.00 Sum

CCM/Ant/DL CRM/Ant/CP 64.67 CCM/Lxh/DL CRM/Lxh/TCRM/Ant/CR 11.82 CRM/Lxh/LCRM/Ant/FA 5.10 CRM/Lxh/FCRM/Ant/IC 0.67 CRM/Lxh/IARM/DLSl/Mix 17.75 ARM/DLSl/Sum 100.00 Sum

CBR Lixhe (Table 6). For the CCM of CBR Lixhe also the SiO 2-sourcewas completely replaced which is not really realistic from a pro- duction point of view because three raw materials are not suffi-cient to manage four metal oxides. As can be noticed in Table 2,the chemical analysis of the fines (ARM/PpF/S9) was significantlydifferent from the material where it was originated from as well as the other porphyry materials that are presented in Fig. 1. This chemical shift could be explained by the fact that hornblende is easier to grind by its lower hardness compared to quartz and feld- spar (Section 2.2) which results in a doubled MgO (wt%) as well as an increased Fe 2O3 (wt%) in the porphyry fines (ARM/PpF/S9) com- pared to the porphyry sand 0/2 (Ag/Pp02Sa/S9). However, it is the lowered SiO 2 (wt%) (Table 2) that results in higher porphyry finesARM (ARM/PpF/S9) dosages in the alternative clinker meal compo- sitions (CCM/Ant,Lxh,Maa/PpF). Furthermore, the porphyry materi- als consist out of higher levels of alkali which will increase the total alkali (wt%) of the final clinker interesting from the point of view of the combined fluxing effect described in Section 3.2.

The preparation of the CCM based on the dolomitic limestone ARM (Ag/DLSl/Mix) had as limiting factors the MgO (wt%) and the ratio between MgO and Fe 2O3 described in Section 3.1. CBR Antoing was ideal to evaluate the influence of high levels of SO 3(wt%) and alkalis (wt%) together with MgO (wt%) approaching the maximum limit of 4.0 wt% and an MgO/Fe 2O3 ratio of 1.4. The alternative clinkers calculated out of the factory of CBR Lixhe will give lower levels of SO 3 (wt%) and alkalis (wt%). Comparing these clinkers with those produced for CBR Antoing will demon- strate their influence on the combined fluxing effect to lower the sintering temperature. Alternative clinkers of ENCI Maastricht with dolomitic limestone ARM would not result in extra informat ion and were therefore after simulation not further investiga ted (Table 6).

4.3. Chemical and TGA analysis

The chemical variation of the porphyry materials was evaluated by plotting them in Fig. 1 by their respective SiO 2 and Al 2O3 (wt%)without LOI since that is the way how the ARM will be fed in reality to the Hot Clinker Meal (HCM). By excluding the LOI, the metal oxi-

Quantity (wt%) CRM + ARM Quantity (wt%)

u 79.44 CCM/Maa/Ref CRM/Maa/Ma 84.38 i 6.63 CRM/Maa/SC 2.90 A 12.34 CRM/Maa/FA 11.02 C 1.59 CRM/Maa/IC 1.70

0.00 ARM 0.00 100.00 Sum 100.00

u 81.27 CCM/Maa/Pp CRM/Maa/Ma 81.09 i 0.00 CRM/Maa/SC 1.95 A 3.07 CRM/Maa/FA 4.10 C 1.57 CRM/Maa/IC 1.43 /S5 14.08 Ag/Pp04Sa/S5 11.42

100.00 Sum 100.00

u 81.55 CCM/Maa/PpD CRM/Maa/Ma 81.11 i 0.54 CRM/Maa/SC 3.13 A 0.00 CRM/Maa/FA 4.53 C 0.71 CRM/Maa/IC 0.95 9 17.20 ARM/PpF/S9 10.29

100.00 Sum 100.00

u 62.53 CCM/Maa/DL CRM/Maa/Ma 77.16 i 7.26 CRM/Maa/SC 5.91 A 10.33 CRM/Maa/FA 10.94 C 1.31 CRM/Maa/IC 1.55 Mix 18.57 ARM/DLSl/Mix 4.44

100.00 Sum 100.00

95

96

97

98

99

100

101

100 170 240 310 380 450 520 590 660 730 800 870 940 1010 1080 1150 1220 1290 1360 1430T (°C)

TGA

(wt%

)

Ag/Pp020FC/S6 Ag/Pp020SC/S9 Ag/Pp04Sa/S5 Ag/Pp02Sa/S9 ARM/PpF/S9

Fig. 5. TGA/DTA analysis of the four selected porphyry materials as well as the porphyry fines ARM.

J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522 517

des, alkali and sulphate are recalculated to 100 wt%. It is clear that the porphyry materials are chemical ly very stable regardles s to their particle size distribution . TGA analyses of the selected por- phyry materials as well as the porphyry fines presented in Fig. 5show two small but distinct mass losses, the first between 450 �Cand 600 �C and a second between 950 �C and 1050 �C. Quite remarkable is the fact that the mass losses in both temperature re- gions are bigger for the fines fraction ARM/PpF/S9, indicating that the mass loss comes most likely from the hornblende that consist partly out of hydroxyl and fluorine. The TGA analysis of the se-

55

60

65

70

75

80

85

90

95

100

105

100 170 240 310 380 450 520 590 660 730T (

TGA

(wt%

)

Ag/DLSl/S11 Ag/DLSl/S5

Fig. 6. TGA/DTA analysis of the three selected dolomitic lim

lected porphyry materials and their chemical analyses (Table 2)show clearly that CaO as well as MgO are not in carbonated form and will not, in contrast with dolomitic limestone, emit CO 2 whenthermally degraded up to 1450 �C. The chemical instability of the dolomitic limestone materials is quite clear out of Fig. 2. This could significantly be ameliorated by selecting them by their particle size. It can also be noticed that CaO is more present in the sand than in coarse aggregates but also that MgO is more concentrated in the sludge than in the sand. The fact that limestone is easier to grind by increasing CaMg(CO3)2 (wt%) explains the higher MgO

800 870 940 1010 1080 1150 1220 1290 1360 1430°C)

Ag/DLSl/S4b ARM/DLSl/Mix

estone as well as the dolomitic limestone mix ARM.

Table 7TGA/DTA evaluation of the reference Cold Clinker Meals of CBR Antoing, CBR Lixhe and ENCI Maastricht.

CCM Anorg CO 2 Total CaO Total MgO Intrinsic Anorg CO 2 CaCO3 MgCO3 Decarb E CaCO 3 Decarb EMat TGA Meas (wt%) XRF Meas (wt%) XRF Meas (wt%) XRF Der (wt%) XRF Der (wt%) XRF Der (wt%) lV s/mg Mat DTA J/g Mat XRF Der

CCM/Ant/Ref 34.0 43.48 1.10 35.32 77.60 2.30 182.3 1390 CCM/Ant/Pp 34.8 43.97 1.18 35.80 78.48 2.48 166.2 1390 CCM/Ant/PpF 34.3 43.68 1.41 35.82 77.96 2.95 166.1 1381 CCM/Ant/DL 35.0 42.66 2.31 36.00 76.14 4.83 162.3 1393 CCM/Lxh/Ref 32.5 45.26 0.88 36.48 80.78 1.84 161.8 1287 CCM/Lxh/Pp 34.03 44.68 1.15 36.32 79.74 2.41 169.7 1305 CCM/Lxh/PpF 34.82 44.45 1.44 36.46 79.33 3.01 154.4 1308 CCM/Lxh/DL 33.8 43.34 2.27 36.49 77.35 4.75 227.7 1316 CCM/Maa/Ref 34.2 44.79 0.97 36.21 79.94 2.03 228.0 1373 CCM/Maa/Pp 33.83 44.62 1.09 36.21 79.64 2.28 219.0 1319 CCM/Maa/PpF 34.01 44.84 1.19 36.49 80.03 2.49 195.2 1319

Table 8Chemical analysis and Bogue calculations of the final clinkers with Porphyry and Porp hyry fines produced in a static kiln.

Clinker Cl/Ant/Pp Cl/Ant/PpF Cl/Lxh/Pp Cl/Lxh/PpF Cl/Maa/Pp Cl/Maa/PpF Cl/Maa/Ref

CaO (wt%) 66.56 66.03 66.49 66.64 65.87 66.79 66.18 SiO 2 (wt%) 22.30 22.65 21.60 21.70 22.60 22.45 21.39 Al 2O3 (wt%) 4.03 4.02 4.32 4.26 4.46 4.47 4.54 Fe 2O3 (wt%) 2.79 2.62 3.93 3.68 3.55 2.75 3.98 K2O (wt%) 0.33 0.45 0.20 0.14 0.24 0.14 0.33 Na 2O (wt%) 0.63 0.51 0.64 0.58 0.59 0.47 0.21 SO 3 (wt%) 0.50 0.63 0.10 0.06 0.04 0.05 0.36 MgO (wt%) 1.80 2.14 1.60 1.90 1.59 1.76 1.52 TiO 2 (wt%) 0.23 0.28 0.27 0.33 0.28 0.31 0.30 P2O5 (wt%) 0.10 0.04 0.18 0.16 0.14 0.15 0.17 Cl (wt%) – 0.12 – – – – –LOI 975 �C (O2) (wt%) 0.43 0.28 0.31 0.22 0.31 0.33 0.48

DoS-factor 45.69 60.50 10.03 6.91 4.14 6.89 65.24 Alite (C3S) 70.37 65.86 71.83 72.44 61.30 67.26 70.61 Belite (C2S) 10.86 15.27 7.75 7.58 18.56 13.63 8.07 Aluminate (C3A) 5.96 6.22 4.80 5.06 5.81 7.19 5.30 Ferrite (C4AF) 8.49 7.97 11.96 11.20 10.80 8.37 12.11

Table 9Chemical analysis and Bogue calculations of the final clinkers of CBR Antoing with dolomitic limestone made in a static kiln.

Clinker Cl/Ant/Ref 1450 �C

Cl/Ant/DL 1450 �C

Cl/Ant/Ref 1400 �C

Cl/Ant/ DL 1400 �C

Cl/Ant/Ref 1350 �C

Cl/Ant/ DL 1350 �C

Cl/Ant/Ref 1300 �C

Cl/Ant/ DL 1300 �C

CaO (wt%) 65.90 64.15 65.94 64.50 65.98 64.26 65.19 64.42 SiO 2 (wt%) 22.27 21.66 22.35 21.84 22.43 21.77 22.35 21.95 Al 2O3 (wt%) 4.14 4.50 4.14 4.14 4.13 4.18 4.14 4.12 Fe 2O3 (wt%) 3.02 2.88 2.97 2.76 2.92 2.77 2.96 2.76 K2O (wt%) 0.59 0.77 0.58 0.83 0.57 0.90 0.76 0.84 Na 2O (wt%) 0.17 0.16 0.18 0.17 0.18 0.15 0.23 0.16 SO 3 (wt%) 0.89 0.86 0.86 0.84 0.83 0.89 1.08 0.81 MgO (wt%) 1.73 3.96 1.75 3.93 1.77 3.93 1.78 3.86 TiO 2 (wt%) 0.25 0.23 0.25 0.22 0.25 0.23 0.25 0.23 P2O5 (wt%) 0.21 0.17 0.2 0.17 0.19 0.17 0.19 0.17 Cl (wt%) Na Na Na Na Na Na Na Na LOI 975 �C (O2) (wt%) 0.48 0.30 0.46 0.30 0.44 0.45 0.73 0.38

DoS-factor 123.42 99.86 119.60 90.80 115.75 92.83 114.51 87.98 Alite (C3S) 66.84 62.13 66.50 64.78 66.16 64.05 63.43 63.75 Belite (C2S) 13.44 15.24 13.92 13.76 14.41 14.10 16.24 14.85 Aluminate (C3A) 5.86 7.05 5.93 6.30 6.01 6.39 5.96 6.25 Ferrite (C4AF) 9.19 8.76 9.04 8.40 8.89 8.43 9.01 8.40

518 J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522

(wt%) in the finer fractions. TGA analyses of the selected dolomitic limestone samples (Fig. 6), show that all CaO (wt%) and MgO (wt%)(Table 3) are present in carbonated form as CaCO 3 and [Ca, Mg] (CO3)2 which will emit CO 2 when sintered in a clinker kiln. The known decarbonation energies found in literature, 1782 kJ/kg for CaCO3 and 1400 kJ/kg for MgCO 3, make it possible to calculate the total decarbonation energy out of the chemical analysis. By its lower decarbonation energy, it is logical that increasing levels

of (CaMg) (CO3)2 in the alternative CCM could lower the energy consump tion of Portland clinker production if the increased MgO (wt%) could replace CaO (wt%) in the mineralogical constituent phases. However, the decarbonation energies (Table 7) out of DTA analysis (lV s/mg) as well as these calculated out of the chem- ical analysis of CRM and ARM (Tables 1–3), the compositi on of the CCM (Table 6) and the known decarbonation energies (J/g) for CaCO3 and MgCO 3 show no big differenc es. The increase of MgO

Table 10 Chemical analysis and Bogue calculations of the final clinkers of CBR Lixhe with dolomitic limestone made in a static kiln.

Clinker Cl/Lxh/Ref 1450 �C

Cl/Lxh/DL 1450 �C

Cl/Lxh/Ref 1400 �C

Cl/Lxh/DL 1400 �C

Cl/Lxh/Ref 1350 �C

Cl/Lxh/DL 1350 �C

Cl/Lxh/Ref 1300 �C

Cl/Lxh/DL 1300 �C

CaO (wt%) 66.28 64.79 66.00 64.26 65.93 64.39 65.89 63.97 SiO 2 (wt%) 21.93 22.12 22.02 22.02 21.89 21.91 21.77 22.23 Al 2O3 (wt%) 4.40 4.26 4.35 4.38 4.42 4.41 4.48 4.18 Fe 2O3 (wt%) 4.21 3.84 4.26 3.80 4.23 3.86 4.20 3.79 K2O (wt%) 0.21 0.11 0.17 0.19 0.20 0.28 0.26 0.36 Na 2O (wt%) 0.20 0.16 0.18 0.18 0.18 0.20 0.20 0.21 SO 3 (wt%) 0.12 0.15 0.08 0.23 0.19 0.30 0.27 0.42 MgO (wt%) 1.28 3.38 1.30 3.44 1.29 3.45 1.35 3.26 TiO 2 (wt%) 0.30 0.31 0.30 0.31 0.31 0.30 0.31 0.31 P2O5 (wt%) 0.24 0.22 0.24 0.23 0.24 0.22 0.24 0.23 Cl (wt%) – – – – – – – –LOI 975 �C (O2)

(wt%)0.39 0.22 0.69 0.56 0.71 0.29 0.63 0.65

DoS-factor 27.47 49.97 21.22 58.37 47.20 60.44 56.32 72.75 Alite (C3S) 67.52 61.48 65.96 59.34 66.24 60.41 66.63 57.92 Belite (C2S) 11.95 17.05 13.38 18.38 12.80 17.52 12.16 20.05 Aluminate (C3A) 4.54 4.79 4.32 5.18 4.56 5.16 4.77 4.67 Ferrite (C4AF) 12.81 11.69 12.96 11.56 12.87 11.75 12.78 11.53

J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522 519

added by the ARM which partly replace CaO coming from the clas- sic limestone source is not sufficient to measure a difference in decarbonation energy. If the limestone CRM could be lowered even more (Table 6) by positive mineralogical influences of the ARM (Section 4.4), a decrease of decarbonation energy could possibly be measure d. Furthermore, the chemical analysis of the final clink- ers presented in Tables 8–10 show that the CCM out of the used ARM and CRM were properly assessed by the simulation program. The increase of MgO in the alternative CCM with porphyry materi- als is moderate but significant contrary to the dolomitic limestone materials were it is quite impressi ve. Based on Fig. 4, the reference as well as alternative clinkers will have an alite formation in the monoclinic form M3 as explained in Section 3.2, which makes the influence of the monoclinic form on the reactivity of alite not rel- evant for this investigation. Also nicely visible in Table 9 is that the reference as well as the alternative clinkers of CBR Antoing have a ratio of Al 2O3/Fe2O3 P 1.38. As explained in Section 3.3, this should result in sufficient liquid formation at 1338 �C. These of CBR Lixhe (Table 10 ) are lower than 1.38. Also the limits described in

Table 11 Mineralogical analysis by XRD of the final clinkers with Porphyry and Porp hyry fines prod

Clinker Cl/Ant/Pp Cl/Ant/PpD Cl/Lxh/Pp

Alite (C3S) (wt%) 73.54 71.04 60.56 Belite (C2S) (wt%) 8.71 12.85 16.56 Aluminate (C3A) (wt%) 9.59 4.75 8.91 Ferrite (C4AF) (wt%) 7.07 10.43 11.21 Free lime (CaO) (wt%) 0.22 0.02 2.08 Periclase (MgO) (wt%) 0.55 0.68 0.46 Arcanite (K2SO4) (wt%) 0.10 – 0.02 Aphthitalite (wt%) 0.16 0.23 0.22

Table 12 Mineralogical analysis by XRD of the reference final clinkers and the clinkers with dolomi

Clinker Cl/Ant/Ref 1450 �C

Cl/Ant/DL 1450 �C

Cl/Ant/Ref 1400 �C

C1

Alite (C3S) (wt%) 64.52 66.08 64.66 6Belite (C2S) (wt%) 19.73 15.01 19.59 1Aluminate (C3A) (wt%) 1.79 5.06 2.73 Ferrite (C4AF) (wt%) 12.86 10.43 11.81 Free lime (CaO) (wt%) 0.23 0.02 0.45 Periclase (MgO) (wt%) 0.39 3.05 0.35 Arcanite (K2SO4) (wt%) 0.32 0.23 0.27 Aphthitalite (wt%) – 0.12 0.02

Section 3.1, MgO (wt%) 6 4.0 wt% and MgO/Fe 2O3 6 1.4 were reached on the final clinkers minimising possible unsoundnes sproblems related to the elevated MgO (wt%).

4.4. XRD analysis

The XRD analyses with Rietveld refinement of the final clinkers presente d in Tables 11–13, show different mineralogical weight percentages than those calculated by Bogue equations out of the chemical analysis of the final clinkers presented in Tables 8–10.Evaluation between the different clinkers could objectively be done by comparing the differences between these theoretical mineralogical Bogue compositions and the real mineralogical XRD compositi ons. A possible fluxing effect could be demon- strated, if a significant increase of the real measure d alite (wt%)occurs in comparison with the theoretical calculated alite (wt%).The goal is to demonst rate if there is a fluxing effect by an increasing MgO (wt%) and if this fluxing effect is related to a com- bined fluxing effect together with R2O (R = Li, Na, K), MgO and SO 3

uced in a static kiln.

Cl/Lxh/PpD Cl/Maa/Pp Cl/Maa/PpD Cl/Maa/Ref

69.91 61.72 66.92 71.33 6.70 16.77 13.84 8.56

11.15 9.21 10.45 4.64 8.57 10.22 7.09 14.89 2.84 0.36 0.43 0.35 0.77 0.62 0.61 0.23 0.27 0.17 0.17 –– 0.34 0.17

tic limestone of CBR Antoing produced in a static kiln.

l/Ant/DL 400 �C

Cl/Ant/Ref 1350 �C

Cl/Ant/DL 1350 �C

Cl/Ant/Ref 1300 �C

Cl/Ant/DL 1300 �C

7.93 64.80 65.69 60.86 59.35 4.90 19.44 17.38 23.35 21.23 4.64 3.67 5.13 5.79 5.69 8.91 10.76 8.00 6.88 6.76 0.07 0.66 0.25 1.34 1.94 3.14 0.31 3.18 0.74 3.16 0.25 0.21 0.23 0.67 0.66 0.15 0.02 0.14 0.14 0.30

55

57

59

61

63

65

67

69

71

73

75

CL/Ant/

PpD

CL/Ant/

DL

CL/Ant/

Pp

CL/Maa

/Ref

CL/Maa

/Pp

CL/Maa

/PpD

CL/Ant/

Ref

CL/Lxh

/DL

CL/Lxh

/Ref

CL/Lxh

/PpD

CL/Lxh

/Pp

Alit

e [w

t%]

0

1

2

3

4

5

MgO

, Nae

q, S

O3

[wt%

]

Alite (XRD) Alite (Bogue) MgO Naeq SO3

Fig. 7. Alite formation sorted from left to right by decreasing mineraliser capacity coming from the combined fluxing effect.

30

35

40

45

50

55

60

65

70

1450°C 1400°C 1350°C 1300°C

wt%

30

35

40

45

50

55

60

65

70

Cl/Ant/Ref BogueCl/Lxh/Ref BogueCl/Ant/DL BogueCl/Lxh/DL BogueCl/Ant/Ref XRDCl/Lxh/Ref XRDCl/Ant/DL XRDCl/Lxh/DL XRD

Fig. 8. Alite (wt%) determined by Bogue and XRD of the classic and alternative clinkers made with dolomitic limestone sludge.

Table 13 Mineralogical analysis by XRD of the reference final clinkers and the clinkers with dolomitic limestone of CBR Lixhe produced in a static kiln.

Clinker Cl/Lxh/Ref 1450 �C

Cl/Lxh/DL 1450 �C

Cl/Lxh/Ref 1400 �C

Cl/Lxh/DL 1400 �C

Cl/Lxh/Ref 1350 �C

Cl/Lxh/DL 1350 �C

Cl/Lxh/Ref 1300 �C

Cl/Lxh/DL 1300 �C

Alite (C3S) (wt%) 65.04 59.50 64.81 54.82 62.78 50.74 60.13 29.02 Belite (C2S) (wt%) 14.93 18.52 15.41 23.58 17.83 27.65 19.16 43.97 Aluminate (C3A) (wt%) 3.68 3.16 3.99 3.60 3.63 2.48 6.43 5.37 Ferrite (C4AF) (wt%) 15.87 16.31 14.82 15.55 14.73 13.60 11.91 12.60 Free lime (CaO) (wt%) 0.23 0.09 0.20 0.15 0.63 0.60 1.89 5.76 Periclase (MgO) (wt%) 0.18 2.16 0.16 2.09 0.11 2.30 0.11 3.02 Arcanite (K2SO4) (wt%) 0.07 0.02 0.05 0.12 0.01 – 0.36 –Aphthitalite (wt%) – 0.16 0.44 0.08 0.17 – – –

520 J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522

J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522 521

as described in Sections 3.2 and 3.3 . The mineralogi cal analyses performed on the alternative clinkers made at 1450 �C are pre- sented in Tables 11–13. In all of the alternative clinkers of CBR Antoing, a significant increase in alite (wt%) of about 2.0–3.0 wt% is distinguisha ble. The increase cannot be explained by an increase of MgO (wt%) built in the alite phase because the in- crease in periclase is bigger than the increase of the total MgO (wt%) (Table 8) compared to the reference clinker. Fig. 7 demon-strates that the increase of alite (wt%) is due to a combined flux-ing effect of MgO, alkali and SO 3. From left to right, the biggest gain up to the biggest decrease in alite formation is plotted. The percentages of MgO, Na eq and SO 3 (wt%) (Tables 8–10) are also in- cluded in Fig. 7. It is clear that on the left of Fig. 7, the clinkers are high in MgO, Na eq and SO 3 (wt%) but even more important a bal- ance is present between alkali and SO 3 (wt%). The clinkers with lower MgO, Na eq or SO 3 (wt%) and where no fluxing effect can be distinguished, are positioned in the middle. The small differ- ences between the real and theoretical alite (wt%) are for these clinkers within the error margin of the XRD measurement and can therefore not be allocated to a possible positive or negative mineralogical effect. At the right, clinkers with high levels (wt%)of alkali but almost no SO 3 (wt%), result in a decrease of alite (wt%) due to the increased viscosity of the melt by the presence of free alkali. The presence of free alkali is demonstrat ed by the small quantities (wt%) of arcanite and aphthitalite in comparis on with the high levels (wt%) of alkali. Although all the CCM were designed to have DoS-factor s between 80 and 120, many of the final clinkers didn’t achieve this goal. The reason of this unbalanc ebetween alkali and SO 3 (wt%) is due to the different volatility of the Cl, SO 3 and alkali in a static lab furnace compare d to a real clinker kiln which was already explained in [3]. On the other hand, it makes it possible to demonst rate the importance of each part of the combined fluxing effect. Based on Fig. 7, it can be con- cluded that a combined fluxing effect [29] exists which can be provoked by the use of the porphyry and dolomitic limestone ARM. The combined fluxing effect is also noticeab le when sinter- ing temperat ure is lowered to 1400 �C and 1350 �C. Both the ref- erence CCM as well as the alternative CCM made with dolomitic limestone ARM of CBR Antoing and CBR Lixhe were sintered at 1450 �C, 1400 �C, 1350 �C and 1300 �C. The resulting alite contents (wt%) measured by XRD are plotted in Fig. 8. The alternative clin- ker made with dolomitic limestone ARM of CBR Lixhe decreases very rapidly in alite (wt%) as of 1400 �C. This trend is not so vis- ible though present in the reference clinker of CBR Lixhe (Fig. 8).In contrary with the clinkers of CBR Lixhe, both clinkers of CBR Antoing maintain the same alite (wt%) up to 1350 �C. As of 1300 �C, all four clinkers decreases in alite (wt%). The difference between the clinkers of CBR Antoing and these of CBR Lixhe, is the presence of alkali and SO 3 (wt%). Although, in the case of the reference clinker of Antoing, an increase in alite (wt%) at 1450 �C was not measured (Fig. 7), there is apparent ly sufficientMgO, alkali and SO 3 (wt%) available to generate the combined fluxing effect resulting in the decrease of the sintering tempera- ture to 1350 �C. This is also the case for the alternativ e clinker made with dolomitic limestone ARM of CBR Antoing which al- ready showed a combined fluxing effect at 1450 �C. Compared to tests performed by Kacimi et al. [24], both the reference as well as the alternative clinkers of CBR Antoing are quite low in free CaO at the different temperatures but it can be remarked that free CaO was significantly lower at 1450, 1400 and 1350 �C when the combined fluxing effect took place. As can be noted out of Table 7,the increase in alite (wt%) formation in the alternativ e clinkers has been obtained without an increase in inorganic CO 2. This means that a more reactive alternative clinker, in the case of CBR Antoing, can be made without an increase of inorganic CO 2;or a clinker with the same reactivity but with less CaO and there-

fore less CO 2 can be made with these higher levels of MgO. Decreasing the CaO by replacing it with SiO 2 will also increase the burnability of the clinker as described in Section 3.3. This should be further investiga ted by evaluating the energy consump- tion up to different temperatures of the reference and alternative clinkers which are composed out of the same final mineralogy (equal alite (wt%)).

5. Conclusion s

By this investigation, it was shown that fines or sludge coming out of the production of porphyry or dolomitic limestone aggre- gates, could be an interesting Alternative Raw Material (ARM) for Portland clinker production. They could, in both cases, have a posi- tive influence on the ecological effects of aggregat es production and application as well as for the excavation of limestone Classic Raw Materials (CRMs) in the case of dolomitic limestone sludge. In contrast with porphyry fines, dolomitic limestone sludge will need a homogenisati on phase to decrease its chemical variation .As could be noted in the different sections, porphyry fines could also replace a CRM with pozzolani c capacity as fly ash. On the con- trary, dolomitic limestone sludge ARM will not replace completely the limestone CRM by its high levels of MgO which will generate an undesired fifth raw material within the Portland clinker produc- tion. The MgO out of dolomitic limestone fines will also be pre- sented to the clinker kiln in carbonated form, resulting in an increase of CO 2 emission with increasing inert periclase formation in spite of the lower decarbonation energy of MgCO 3. Furthermore it was demonst rated that MgO when cautiously introduced by the investiga ted ARM in line with well defined limitations, should rather be regarded for its positive mineralogical influence on clin- ker than avoided because of its possible periclase formatio n and re- lated unsoundnes s propertie s. MgO combined in combination with alkali and SO 3 has fluxing activities whereby the alite formation is promote d which could improve the burnability of the alternative clinkers. Furthermor e, it was also demonstrat ed that in combina- tion with alkali and SO 3, a decrease of the liquid formation temper- ature is noticeable by which equal mineralogi cal composition can be attained at 1350 �C. Lowering the sintering temperature to 1350 �C could safely be considered in combination with the in- creased alite formation at 1450 �C. The combined fluxing effect is not as distinct as this of CaF 2 but at the other hand easier and cheaper to introduce in a clinker kiln by the ARM. It is also a fact that it is harder for clinker producers to attain too low than suffi-cient SO 3 (wt%) and alkali (wt%) to profit from a combined fluxingeffect. Therefore the fine tuning of MgO (wt%) up to 2.0–2.5 wt% in Portland clinker could be recomme nded to fully benefit the miner- alogical advantages. Further investigatio n should be performed to evaluate physical properties of cement based on these ARM. Addi- tional to materials with decarbonate d CaO, ARM who consists partly out of decarbonated MgO could act as a valuable raw mate- rials for Portland clinker production. Porphyry and to a lesser ex- tent dolomitic limestone fines or sludge should therefore, together with the already available alternativ e fuels and raw mate- rials [26], be considered as a way to get in line with the Cement Sustainabi lity Initiative.

Acknowled gments

The authors wish to thank the Central lab of Sagrex Quenast and the cement research lab of ENCI Maastricht for their support, Els Bruneel for her aid during the practical execution of the tests, Jo Lejeune for his help during the XRD tests and Eleni Arvaniti for crit- ically reading the manuscript.

522 J. Schoon et al. / Construction and Building Materials 43 (2013) 511–522

References

[1] Damtoft JS, Lukasik J, Herfort D, Sorrentino D, Gartner EM. Sustainable development and climate change initiatives. Cem Concr Res 2008;38: 115–27.

[2] Mikulcic H, Vujanovic M, Duic N. Reducing the CO 2 emissions in Croatia’s cement industry – the precalciner model. In: 6th Dubrovnik conference on sustainable development of energy, water and environment systems; 2011.

[3] Schoon J, Van der Heyden L, Eloy P, Gaigneux EM, De Buysser K, Van Driessche I, et al. Waste fibre cement: an interesting alternative raw material for asustainable Portland clinker production. Constr Build Mater 2012;36:391–403.

[4] Puertas F, García-Díaz I, Barba A, Gazulla MF, Palacios M, Gómez MP, et al. Ceramic wastes as alternative raw materials for Portland cement clinker production. Cem Concr Compos 2008;30:798–805.

[5] Pardo N, Moya JA, Mercier A. Prospective on the energy efficiency and CO 2emissions in the EU cement industry. Energy 2011;36:3244–54.

[6] NBN EN 1097: Essais pour determiner les caractéristiques mécaniques et physiques de granulats; 1996.

[7] Hojamberdiev M, Muhamedbaeva Z, Madhusoodana CD. Use of natural and thermally activated porphyrite in cement production. Constr Build Mater 2009;23:2757–62.

[8] Tsivilis S, Chaniotakis E, Badogiannis E, Pahoulas G, Ilias A. A study on the parameters affecting the properties of Portland limestone cements. Cem Concr Compos 1999;21:107–16.

[9] EN 197-1: Cement – Part 1: composition, specifications and conformity criteria for common cements; 2000.

[10] EN 197-2: Cement – Part 2: conformity evaluation; 2000. [11] Taylor HFW. Cement chemistry, 2nd ed., vol. xviii. London: T. Telford; 1997. p.

459.[12] Javed I. Bhatty: role of minor elements in cement manufacture and use.

PCA599 Research and Development Bulletin RD109T. Skokie, Illinois, U.S.A.: Portland Cement Association; 1995.

[13] Waldemar A. Klemm: cement soundness and the autoclave expansion test – an update of the literature. PCA R&D Serial No. 2651. Skokie, Illinois, USA: Portland Cement Association; 2005. 20 p.

[14] ASTM C 150: Standard specification for Portland cement; 2000. [15] NZS 3122 (A1): Specification for Portland and blended cements; 2009. [16] AS 3972: General purpose and blended cements; 2010.

[17] Mo Liwu, Deng Min, Tang Mingshu. Effects of calcination condition on expansion property of MgO-type expansive agent used in cement-based materials. Cem Concr Res 2010;40:437–46.

[18] Chatterji S. Mechanism of expansion of concrete due to the presence of dead- burnt CaO and MgO. Cem Concr Res 1995;25(1):51–6.

[19] Akin Altun I. Effect of CaF 2 and MgO on sintering of cement clinker. Cem Concr Res 1999;29:1847–50.

[20] Katyal NK, Ahluwalia SC, Parkash R, Sharma RN. Rapid estimation of free magnesia in OPC clinker and 3CaO:1SiO 2 system by complexometry. Cem Concr Res 1998;28:481–5.

[21] Standard EN 196-3: Determination of setting times and soundness; 2009. [22] Iwao Maki, Koichiro Fukuda, Hideto Yoshida, Junichi Kumaki. Effect of MgO

and SO 3 on the impurity concentration in alite in Portland cement clinker. J Am Ceram Soc 1992;75(16):3163–5.

[23] Maki I, Goto K. Factors influencing the phase constitution of alite in Portland cement clinker. Cem Concr Res 1982;12:301–8.

[24] Larbi Kacimi, Angélique Simon-Masseron, Abdelhamid Ghomari, Zoubir Derriche. Influence of NaF, KF and CaF 2 addition on the clinker burning temperature and its properties. CR Chimie 2006;9:154–63.

[25] Blanco-Varela MT, Palomo A, Vazquez T. Effect of fluorspar on the formation of clinker phases. Cem Concr Res 1984;14:397–406.

[26] 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.

[27] Surana MS, Joshi SN. Use of mineralizers and fluxes for improved clinkerization and conservation of energy. Zement-Kalk-Gips 1990;43:43–7.

[28] Klemm WA, Jawed I, Holub KJ. Effects of calcium fluorided mineralization on silicates and melt formation in Portland cement clinker. Cem Concr Res 1979;9:489–96.

[29] Ghosh SN. Part 1, Improving energy efficiency in Portland clinker. Cem Concr Sci Technol;1:191,Thomas Telford; 1991.

[30] Johansen Vagn, Christensen NH, Smidth FL. Rate of formation of C3S in the system CaO–SiO2–Al2O3–Fe2O3–MgO with addition of CaF 2. Cem Concr Res 1979;9:1–6.

[31] EN 13043: Aggregates for bituminous mixtures and surface treatments for roads, airfields and other trafficked areas; 2002.

[32] EN 12620: Aggregates for concrete; 2002. [33] NBN EN 206-1: Concrete – Part 1: specification, performance, production and

conformity; 2001.