MINERALISED CLINKER AND CEMENT S Rasmussen D Herfort Aalborg Portland Denmark A B S T R A C T . Mineralised clinkers and the potential benefits of using cements based on mineralised clinkers have been studied for many years. During the past two decades attention has mainly focused on the benefits arising from the combined use of fluorine and S 0 . Until recently, however, no producers of grey cement, have utilised this technology. The Danish producer, Aalborg Portland has, since 1994, produced a cement based on mineralised clinker as one of its main products for the domestic market. The clinker contains approximately 2% S0 and 0.2% F. The combined mineralising action lowers burning temperatures in the kiln by approximately 100°C, giving considerable reductions in emissions of NO in particular. Cements ground from this clinker are more reactive with both higher early and 28 day strengths than for otherwise identical non-mineralised cements. Where rapid strength devel opment is not desirable this technology can be used to partially replace the clinker by a mate rial such as limestone. In Denmark this is achieved by replacing 14% of the clinker with a limestone rich dust giving a further reduction in NO emission in addition to a considerable reduction in C 0 emission by weight of the cement produced. 3 3 x x 2 Keywords: Mineralised cement, Clinker, Fluorine, Sulphate, Limestone filler. M r Soren Rasmussen is Manager of the Project Service Department at Aalborg Portland. His main interests include the properties of mineralised and composite cements, and the ef fects of admixtures. M r D u n c a n Herfort is chief geologist at Aalborg Portland. His responsibilities include all mineralogical aspects associated with raw material assessment, process and quality control, and concrete analysis. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 60 Rasmussen, Herfort INTRODUCTION During recent decades the cement industry has had to face growing demands as a result of in creasing fuel costs, stricter emission limits, higher expectations regarding cement quality, and in many cases increased competition. Reduced fuel consumption has mainly been achieved by investing in more efficient kilns. Modern precalciner kilns, for example, use only half the amount of fuel required by their wet process predecessors to produce a tonne of clinker, and at much higher outputs. Further reductions in fuel consumption have been achieved by par tially replacing the clinker with pulverised fly ash, ground blast furnace slag, natural pozzo lans, limestone dust etc. The wide range of such composite cements produced in Europe is exemplified by the long list of cement types specified in the European pre-standard, ENV 197. Another way of reducing fuel consumption as well as emissions of C 0 and NO is by using a mineraliser where this is not already present in the raw materials. The mineralisers which have received most attention, particularly since the work done by Blue Circle in the 1970's and early 1980's [1,2], are CaF and CaS0 , because of the marked improvements in clinker combinability and strength development of the cement when these components are used in combination. This technology has been successfully applied at Aalborg Portland since 1994, with considerable savings in fuel consumption and improvement in cement quality . 2 2 x 4 MINERALISED CLINKER Although at times used loosely, the term mineralisation in clinker production is defined in the same way as it is in geology, i.e. it is the process of promoting mineral formation. A miner aliser is, therefore, the agent by which this is achieved, for example by lowering the tem perature of formation of a mineral through solid solution, or by increasing the content of the melt, or changing the properties of the melt. Strictly speaking, then, a fluxing agent is also a mineraliser if it promotes mineral formation. Not all fluxing agents are mineralisers, how ever, since some are known to inhibit the formation of clinker minerals. The term is conven tionally used to describe the minor components which promote the formation of alite. A1 0 and F e 0 are, therefore, not regarded as mineralisers in the production of Portland cement clinker, because they occur as major components. For a comprehensive review of mineralis ers in Portland cement clinker, the reader is referred to Moir & Glasser . Some of the ma jor trends are briefly described below. 2 2 3 3 The effects that transition metals have on the formation of alite varies considerably with no clear relationship with their effect on the properties of the melt. Of the P-block elements, S and P are known to inhibit the formation of alite, except where this is offset by sufficiently high contents of other minor components, particularly F. The presence of halogens, notably F and CI, are known to promote the formation of alite (or alinite in the case of CI). The effect of the alkalis is strongly dependent on the degree of sulphurisation. The above minor components are always present in cement but, apart from the alkalis and sulphates, are usually at too low concentrations to have any significant effect. A combination of sulphate and fluoride is known to have a particularly positive effect, although it is unusual for the desired combination to occur naturally in the raw materials and/or fuel. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Mineralised Clinker and Cement 61 A small number of producers, including Aalborg Portland have utilised this technology through the controlled addition of CaS0 and CaF to the raw mix used to produce grey clinker. 4 2 Two types of mineralised clinker are currently produced at Aalborg, both in its 5500 t/d semiwet, 2-stage separate line precalciner kiln, the details of which are published elsewhere [5,6]. The composition of the more normal clinker is given in Table 1, with S 0 and F contents of 1.9% and 0.18%, respectively. A high C S clinker is also produced in which the F content is increased to 0.23% as shown in Table 2. Combination to between 1 and 2% free lime is achieved at approximately 1350°C which is about 100°C below that for the non-mineralised clinker. The bulk of the sulphate melt forms between 1000°C and 1200°C. Most of the alite forms fairly rapidly over a narrow temperature interval between 1150°C and 1200°C before any significant oxide melt formation begins. The rapid, early formation of alite is partly due to its greater stability at lower temperatures, and partly to the high rate of diffusion of Ca through the low viscosity sulphate melt. About 2/3 of the sulphate occurs in a readily water soluble form, mainly as Ca-langbeinite, with the remainder incorporated in the silicate phases, with the highest concentrations occurring in the belite phase which significantly in creases its hydraulic reactivity. 3 3 2+ Low temperatures for clinker formation and relatively high partial pressures of 0 prevent ex cessive volatilisation of the sulphates. Volatilisation of F is negligible. Ring formation and other volatile related kiln build-ups are therefore limited. Direct fuel savings are marginal at approximately 5%. Formation of NO in the kiln is reduced by up to 50%. Since the total NO emission from this type of kiln is also dependent on fuel derived NO formed in the calciners, the overall reduction in emission of NO can vary from 10 to 30% depending on the nitrogen content of the fuel. 2 x x x x MINERALISED CEMENT The properties of cement based on all known mineralisers is beyond the scope of this paper and only mineralised cement prepared from clinker containing controlled levels of S 0 and F will be discussed. The properties of mineralised cement are no different from those of normal Portland cement in terms of the dependency on the relative content of clinker minerals, fine ness, ratio of soluble sulphate to C A, etc. The only difference lies in the modified reactivity of the silicate phases which gives longer setting times and faster strength development. 3 3 Setting Times Setting times are increased with increased contents of F in the alite phase. Work by Renichi Kondo et al  indicates that this is due to the precipitation of CaF on the surface of the alite crystals during initial hydration. The relationship between setting time and the fluorine con tent of the mineralised cement produced at Aalborg shown in Figure 1 corresponds to 250 minutes/%F. The setting times were determined from pastes prepared to normal consistency, according to EN-196, and do not necessarily correspond to the setting times of concretes which often contain water reducing and air entraining additives. 2 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 62 Rasmussen, Herfort Studies at Aalborg Portland have shown that the difference in concrete setting times is ap proximately 30 minutes for the cements shown in Table 1 regardless of the type and dosage of the admixture. In other words, even though the admixture may increase the setting time of the concrete by up to 5 times that of the paste, the difference between concretes containing the mineralised and non-mineralised cements remains constant at approximately 30 minutes. In relative terms, then, the effect on concrete is not as great as would appear from the EN-196 setting times alone. 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18 0.20 C O N T E N T OF FLUORINE IN C E M E N T , % Figure 1 Relationship between initial setting time (EN-196) and content of fluorine in cement Compressive Strengths Apart from the benefits on clinker production discussed above, the main advantage of miner alisation lies in the increased reactivity of the cement resulting in a more rapid strength de velopment. The increased reactivity is the result of increased solid solution of the mineralis ers, i.e. S and F in the silicate phases, along with an additional incorporation of Al which this facilitates. An example of the enhanced strength development of mineralised cement com pared to a similar, but non-mineralised cement, is shown in Figure 2. The strengths plotted are the average of two, and sometimes three separate EN-196 determinations. The chemical and physical characteristics of these cements are shown in Table 1. The higher 7 and 28 day strengths are mainly due to the increased reactivity of the belite phase in the mineralised clinker. With the possible exception of the ferrite phase, which makes little contribution to strength development, belite is normally the least reactive phase, making relatively little contribution to 28 days strengths. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Mineralised Clinker and Cement 63 In mineralised clinker, however, a far higher proportion of the belite has reacted at 28 days, to the extent that increasing the ratio of C S to C S in the clinker (essentially lowering the LSF) results in comparable or often higher 28 days strength. 2 3 80 10 I 1 1— 1 1 1—i 10 • • i • i 1 100 TIME, days Figure 2 Compressive strength of mortars with mineralised cement no MI and nonmineralised cement no NO Table 1 Clinker and cement data NON-MINERALISED CEMENT, no. NO MINERALISED CEMENT, no. MI CLINKER: C S° C S'> CA C AF Free lime Fluorine S 0 in clinker N a 0 eqv. acid sol. % % % % % % % % 52.9 20.1 6.1 11.6 1.36 0.06 0.90 0.59 54.1 20.0 6.2 10.8 0.98 0.18 1.90 0.56 CEMENT: Fineness (Blaine) 45 urn residue S 0 in cement Limestone extender m /kg 2 426 2.4 2.94 4.0 412 2.5 2.91 4.0 3 2 3 4 3 2 3 % % % Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 64 Rasmussen, Herfort 1) Bogue composition corrected for S 0 and free lime Although the early strengths, i.e. 1 and 2 day strengths, are essentially the same for the two cements shown in Figure 2, higher early strengths can be achieved by increasing the LSF of the mineralised clinker which is readily accomplished owing to its easier burnability. Strength development for this type of cement is shown in Figure 3 for the cements labelled MH420 and MH251 (ground to specific surface areas of 420 and 251 m /kg), with relevant chemical and physical data shown in Table 2. The results show, among other things, that very high early strengths can be achieved without grinding to excessive fineness. This allows the addition of a relatively inactive filler such as limestone . The strength development obtained with a replacement of the clinker by 14% calcareous filler (cements MH420L and MH251L) are also plotted in figure 3. The cement labelled MH420L in Table 2 has been produced in Denmark since 1994, where it is the preferred cement for the pre-cast and con crete product producers. It is formally designated by ENV 197-1 as a PC 52.5 R, Portland limestone cement (type II/A-L). 3 2 For the cements to comply with the upper limit of 62 MPa for ENV 197, class 42.5 cements, the clinker fineness must be kept below 250 m /kg as shown by the strength plots in Figure 3. The presence of a relatively high content of fine filler (MH251L) eliminates the risk of bleeding which would normally be a problem at such a low fineness of the clinker. Com pared with conventional PC 42.5 cement, savings achieved on grinding would be at least 20%. 2 For both filler cements reductions in fuel consumption are more or less the same as the level of clinker replacement, i.e. 14% in the examples shown. Reductions in NOx emissions are estimated at slightly more than 30%. 80 , 10 I 1 , 1 1 1 1 10 ' ' 100 TIME, days Figure 3 Compressive strength of mineralised cement with relatively high content of C S and with varied fineness and content of limestone filler, mortar strength determined according to EN-196 3 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Mineralised Clinker and Cement 65 Table 2 Clinker and cement data. MINERALISED CEMENT MH251 CLINKER C S'> CS CA C AF Free lime Fluorine S 0 in clinker N a 0 eqv. acid sol. 3 ]) 2 3 4 3 MH251L MH420 MH420L 66.4 6.7 7.9 9.7 1.80 0.23 1.90 0.59 % % % % % % % % 2 CEMENT S 0 in cement Fineness of clinker (Blaine) 45 um residue Limestone filler 3 % 2 m /kg % % 3.01 251 19.0 0.0 2.92 251 17.3 14.0 3.04 420 2.1 0.0 2.95 420 2.0 14.0 1) Bogue composition corrected for S 0 and free lime 3 Durability Experience in Denmark to date has been that concrete produced from mineralised cement is equally durable to concrete produced from conventional Portland cements. This is docu mented in a report by the Danish Institute of Testing of Building Materials (DTI) , and in publications by the present authors [10,11]. Despite having many properties in common with the other halogens, including chlorine, fluo rine does not cause corrosion of steel reinforcement due to the low solubility of CaF in pore solutions found in concrete . 2 Another concern has been that clinker containing high sulphate contents may give rise to a type of secondary ettringite formation which may lead to deleterious expansion in much the same way as delayed ettringite formation (DEF). As noted in the section on clinker, however, most of the sulphate present in the clinker is in a readily soluble form which contributes to regulation of the initial set. All but a negligible fraction of the remaining sulphate, which occurs in the unreacted sili cates, is available for any further ettringite formation, and even here an invariable excess of A1 0 precludes any ettringite formation whatsoever. Numerous experimental studies [10,13,14] have recently shown that expansion is unaffected by the level of sulphate in the clinker. 2 3 Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. 66 R a s m u s s e n , H e r f o r t CONCLUSIONS The production of mineralised clinker using controlled additions of fluorine and sulphates lowers burning zone temperatures by approximately 100°C. Although this results in fuel savings of only 5%, NO formed in the kiln is reduced by up to half, which in a precalciner kiln typically equates to an overall reduction in NO emission of approximately 20%. x x In terms of fuel consumption and C 0 emission the main advantage of producing mineralised cement lies in its greater hydraulic reactivity which allows the replacement of clinker by C 0 neutral additions such as limestone. 2 2 REFERENCES 1. MOIR, G, K. Mineralised High Alite Cements. World Cement, 1982, Vol. 13, No. 10, pp 374-382. 2. MOIR, G, K. Improvements in the early strength properties of Portland cement. Phil Trans. Lond., 1983, A310, pp 127-138. 3. BORGHOLM, H, E, HERFORT, D, AND RASMUSSEN, S. A New Blended Ce ment Based on Mineralised Clinker. World Cement, 1995, Vol. 8, pp 27-33. 4. MOIR, G, K. Effect of the Use of Mineralisers, Modifiers and Activators. 9th ICCC, New Delhi, 1998, Vol. l,pp 125-152. 5. BORGHOLM, H, E, AND NIELSEN, P,B. Ein neues Halbtrocken-Ofen-System fur 4000t/d im Aalborg Portland-Zementwerk. ZKG Inernatioanl, 1988, Vol. 41, No. 12, pp 595-600. 6. BORGHOLM, H, E. Comissioning the world's largest semi-dry process kiln system. World Cement, 1989, Vol. 20, No. 3, pp 72-79. 7. KONDO, R, DAIMON, M, SAKAI, E AND USHIYAMA, H. The Influence of Inor ganic Salts on the Hydration of Tricalcium Silicate. J. appl. Chem. Biotechnol. 1977, 27, pp 191-197. 8. BORGHOLM, H,E, AND DAMTOFT, J, S. EUROPEAN PATENT, EP 0640 062 B l , 1993. 9. PETERSEN, E, J, AND HAUGAARD, M. New Cement Types: Evaluation of the Suitability of Basis-Cement (= Portland Limestone Cement Based on Mineralised Clinker) in Concrete designed for Moderate and Aggressive Environments. DTI Byggeteknisk Institut, March 21, 1994, In Danish. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved. Mineralised Clinker and Cement 67 10. HERFORT, D, RASMUSSEN, S, JONS, E, AND OSB^CK, B. Mineralogy and Performance of Cement Based on High Sulphate Clinker. ASTM Symposium on In ternal Sulphate Attack on Cementitious Systems: Implications for Standards. In press, Cement, Concrete and Aggregates. 11. HERFORT, D, S0RENSEN, J AND COULTHARD, E. Mineralogy of Sulphate Rich Clinker and the Potential for Internal Sulphate Attack. World Cement, 1997, Vol. 28, No. 5,pp 77-85. 12. ESCUDERO, M, L, AND MACIAS, A. Corrosion of Reinforcing Steel in Mortar of Cement with CaF as a Minor Component. Cement and Concrete Research, 1995, Vol. 25, No. 2, pp 376-386. 2 13. KELHAM, S. Effects of Cement Composition and Hydration on Volume Stability of Mortar. 10th ICCC, Gothenburg, 1997, Vol. 4, paper iv060. 14. TENNIS, S, BHATTACHARJA, S. AND KLEMM, W. Ambient Temperature De layed Ettringite: A Reapraisal. ASTM Symposium on Internal Sulphate Attack on Cementitious Systems: Implications for Standards. In press, Cement, Concrete and Aggregates. Downloaded by [ Griffith University] on [25/10/17]. Copyright © ICE Publishing, all rights reserved.