Clinker C3S paramount

Dr Clark explores how process knowledge and particularly the maximisation of C3S content in clinker together with that C3S hydraulic reactivity are the key to a successful blended cement strategy.

During the recent Cemtech Middle East & Africa event in Marrakech, Morocco in mid-February, the moderator delivered a workshop on the importance of process knowledge for product development in the cement industry. Product development in the cement industry takes a number of forms with the manufacturing and marketing of blended cements being by far the most widespread.

In many countries the manufacture and supply of blended cements has become the norm with far greater volumes of these products being sold than pure Portland cements. Blended cements are Portland cements where a significant proportion of the clinker in cement is replaced by supplementary cementitious materials such as ground granulated blastfurnace slag (GGBS), pulverised coal fly ash (PFA), silica fume, natural pozzolans or limestone. There are other supplementary materials that can replace clinker in blended cements.

The manufacture and sale of blended cements has become increasingly widespread for a number of reasons:
• production capacity, sales, revenue and profits of a cement company are swelled
• carbon intensity of cement is reduced
• blended cements are genuinely better that pure Portland cements in a number of mortar and concrete applications.

The greater the replacement of clinker by supplementary materials the greater is the amount by which the capacity and sales of a cement company are swelled and the greater the reduction in the carbon footprint of the cement.

Meeting strength challenges

However, any cement company pursuing this strategy has to recognise that there is one major drawback encountered with blended cements. That is the lower compressive strength exhibited by mortar and concretes produced with these cements at one, two, three and seven days. These lower early compressive strengths mean that the time required before demoulding or the removal of concrete formwork is lengthened. This raises serious issues in block or precast concrete manufacture and also in some construction applications. A successful blended cement strategy has to recognise and minimise this drawback. A way has to be found to maximise the clinker replacement level while minimising the reduction in early compressive strengths.

The reasons for the reduction in early compressive strength are easy to understand. The supplementary materials may exhibit cementitious properties themselves, contributing to the ultimate strength of mortar or concrete, but in all cases the cementing reaction is slower than with Portland cement clinker. Significant contributions by these materials to strength development are not apparent until beyond seven days from mixing. In terms of the early strength of the mortar or concrete the supplementary materials are simply diluting the clinker in the cement.

Maximising C3S content or hydraulic reactivity

How then can the clinker replacement level be increased while this dilution of the clinker’s early strength-giving properties is minimised? To identify the way to do that an understanding of the hydration of the Portland cement clinker is required. Essentially the early strength development of the clinker derives from the hydration of the C3S in the clinker. It is the dilution of the C3S that leads to the lower early strength development of blended cements. Therefore this effect can be minimised by maximising the C3S content or maximising the hydraulic reactivity of that C3S. It is this maximisation of the C3S content of clinker and maximising the reactivity of that C3S that requires process knowledge.

Increasing C3S content

Achieving a high C3S content of clinker might be thought to be relatively simple. Clinker C3S contents of up to ±67 per cent in grey cement clinker can be achieved by raising the lime saturation of the clinker towards 100 per cent. Beyond 100 per cent lime saturation the additional lime will be present as free lime and there will be no additional C3S. That is easy enough for any cement chemist to do, albeit at the cost of increased fuel consumption on the kiln as the clinker becomes progressively more difficult to fully combine with rising lime saturation. These are indeed the first steps, but there are many more subtle ways to boost the C3S content of clinker and the reactivity of that C3S.

Firstly, we need to recognise that industrial Portland cement clinker is not an equilibrium product with a mineral composition as given by the Bogue calculations. The rapid cooling of the clinker as it passes under the flame in the kiln and falls into the cooler results in non-equilibrium cooling with the liquid flux in the clinker being frozen into a glass. This non-equilibrium cooling means that the clinker mineral composition is frozen at its high-temperature equilibrium with significant deviation from the Bogue composition. With alumina modulus higher than 1.7 there is more C3S present than indicated by the Bogue calculation. The higher the alumina modulus above 1.7, the greater is the increase in the C3S content above the Bogue calculation. From this we can infer that one route to maximising the C3S content of clinker is to maximise the alumina modulus of the clinker. The moderator knows that this has the desired effect of promoting early strength development of cement from long experience with high alumina modulus clinkers and cements.

Conversely, if the alumina modulus is below 0.9 then the amount of C3S present at the sintering temperature and in the industrially-cooled clinker will be lower than the Bogue calculation. You certainly do not want that to be the case if the intention is to produce blended cements with the clinker. (There would be no point anyway. The only reason to lower alumina modulus to 0.9 would be to produce a sulphate-resisting clinker and a blended cement will provide enhanced sulphate resistance).

The next consideration is to guard again any other factors that would reduce the amount of C3S in the clinker. Species to be wary of are sulphate in excess of alkalis, potassium oxide itself and phosphates. Sulphate in stoichiometric excess of alkalis in clinker will be present as calcium sulphate, CaSO4. That ties lime up in calcium sulphate, reducing the effective lime saturation and the potential C3S content of clinker. (There is an added problem that calcium sulphate may breakdown to lime and sulphur dioxide at a late stage in the clinkering process resulting in high free lime clinker).

High alkali content of clinker acts as an inhibitor to C3S formation in the kiln. High potassium oxide in excess of sulphate is particularly problematic as a modified C2S mineral is formed which cannot be converted to C3S. This can effectively block the formation of C3S in the kiln, resulting in high free lime, low C3S clinker which is particularly unsuited to the manufacture of blended cements.

The final species to be wary of in this regard is phosphate, P2O5. Phosphate above 0.5 per cent in clinker leads to progressive decomposition of C3S to form a solid solution of tricalcium phosphate and C2S. Again to be avoided at all costs if the clinker is to be used for the manufacture of blended cements. With Cemtech being in Morocco this is a particular warning for cement manufacturers in that country with the largest phosphate reserves in the world.

Are there species that will lead to an increase in the C3S content of clinker beyond that indicated by the Bogue calculation? Fluoride is the species of interest in this regard as it causes a widening of the primary phase field of C3S and allows clinkers with lime saturation up to 104 per cent to be fully combined with significantly higher C3S content. This is the means by which Aalborg produce 52.5-grade blended cements in Denmark.

Improving C3S hydraulic reactivity

So high alumina modulus, high lime saturation factor and fluoride mineralisation are the routes to high C3S content and the production of clinkers that will allow the maximum replacement of that clinker in blended cements. What about the hydraulic reactivity of that C3S?

C3S can be present in triclinic, monoclinic or rhombohedral crystal forms in clinker. Fluoride is additionally effective for producing reactive C3S because it promotes the presence of the C3S in the hydraulically-reactive rhombohedral form. The other means to producing C3S in the more reactive mineral forms is to achieve a sharp flame and temperature profile in the kiln by means of main burner design and adjustment.

So the presence of fluoride and a sharp flame result in the C3S being in the best mineral form for hydraulic reactivity. Conversely a long, lazy flame will be bad for C3S reactivity. Is anything else bad for C3S reactivity? The species to be wary of here is MgO. MgO will enter solid solution in the clinker minerals substituting for CaO, and thereby raising the lime saturation somewhat. However, there is some evidence that MgO can promote the C3S in clinker being in one of the less reactive monoclinic crystal forms and that this can severely reduce early strength development of cement made from the clinker. In one case the MgO content proved to be the most strongly negatively-correlated species with early strength of cement.

The foregoing demonstrates that there is more to the production and marketing of blended cements than initially meets the eye. Process knowledge is certainly required to get the best from a blended cement strategy!

Article first published in International Cement Review, March 2013.