Arthur M Harrisson Ltd

Quality                                                                Contact

Home   Background     Raw Materials    Limestone      Chalk     Cement Chemistry      Secondary raw Materials     Clinker Microscopy  Solutions

What do I mean by "Quality"?

There are as many definitions of good quality cement as there are uses for cement.  Early strength, late strength, durability, water reduction properties, colour.

By applying simple data analysis techniques combined with years of experience of the materials and of the process I have faciliated improvement to the quality of cement for the customers at each plant with which I have been associated.  The same approach applies to whichever aspect of quality is of interest to a plant. The chart to the right shows the change over four years at two UK cement plants which had been poor performers with fly ash in concrete. Through all this period the average mortar 28 day strength remained close to 60 N/mm² as required to maintain the cements' designations.

While I am familiar with the cement standards used in many countries, I am not an expert in the fine detail of this and am not a Quality Assurance  specialist, rather I am very experienced in understanding what the customer requires and providing the best product for the desired purpose.

What can be done?

I offer an analysis of plant materials, data and products to help to achieve improvements in the market.  Although a visit to site is sometimes helpful, the analysis is laboratory based and thus avoids the necessity for expensive travel and accommodation while being applicable anywhere in the World.  

It is unlikely to be possible to produce a different cement to suit each customer and it is therefore generally necessary to optimise the properties of the cement for the plant's main customer base while also making what improvements are possible in other areas without adversely affecting the primary targets.
Prediction of Quality

Quality prediction is frequently available within  cement companies based on methods of variable robustness.  Customers may be aware of the predictions, salespeople use them or not depending on the perceived value of the figures.

The cement makers are often not aware of the predictions, nor of what aspects of the product most influence the particular properties required by each customer.  Quality is one thing to the cement plant - usually conformity to the local standards.  To a precast manufacturer it will usually be the early strength in concrete (not necessarily in mortar prisms) which controls the company's profit levels.  

To ready mix concrete makers the quality of the cement is overwhelmingly dependent on the water demand of the freshly mixed concrete.  If fly ash is being used the perception of quality will not be the same as for plain Portland cement concrete.  If blastfurnace slag is being used it will be different again.

Figure 1 shows that relationships between some properties and quality are not simple.  If the alkali level of a cement increases it will be advantageous in ready mix concrete if the clinker sulphate is high but disadvantageous if the clinker sulphate is low.  This is because alkali will combine with sulphate in the clinker to provide a very soluble source of sulphate to retard gypsum reaction and improve workability thus reducing water demand but, if there is insufficient sulphate present, the alkali will become incorporated in the C₃A, affecting its reactivity with unpredictable effects.

It is always useful for customers if the cement provider can give a prediction of cement quality immediately after production instead of waiting 7 or 28 days to see how the test results turn out.  It is also invaluable to the cement maker to understand why the cement is not performing as expected.  There are generally about 5 or 6 factors which control how a cement will behave in practice but the effects of varying these factors will not always be the same and the history of an individual plant must be analysed to obtain the most accurate predictions for a given parameter.  The factors may include fineness, C₃S content, alkali level and SO₃ content, free lime and loss on ignition (LOI). The LOI is largely determined in most cements by the quantity of limestone filler added at the milling stage, the loss at high temperature being due to calcination of CaCO₃.  It can also be a result of hydration of the clinker before milling, perhaps as a result of being stored outside in a damp climate.  While the limestone can have a beneficial effect in assisting water demand in concrete, the effect of hydration of cement crystals is generally negative and if this is a likely event it is therefore helpful if the two causes of loss can be separated.

Having identified the primary causes of change in performance those which cause the most variability are tabulated, together with the performance parameter being studied, and a multiple linear regression analysis is performed.  This produces an equation reflecting the influence of each variable on the performance.  The example below is for 28 day strength in mortar prisms

        28d strength =  64-(0.4*45µ)+(0.2*C₃S)-(13*Na₂O_e)-(2*LOI)

So increasing the 45µ residue decreases strength, increasing C₃S increases strength and increasing alkali or loss will give a decrease in 28 day mortar strength.

While for 28 day mortar strength at a fixed water content these figures are fairly typical, of far more significance in ready mix concrete is the quantity of water which is required to obtain a specified workability.  A major influence is the rapidity with which sulphate can be put into solution with in mixing water and thus retard the flash set of C₃A.  The rapidity is in turn affected  by how much sulphate is present as alkali sulphate in the clinker and how much dehydration of gypsum has taken place.  A second influence is the particle size distribution of the cement grains.  A high efficiency separator on a cement mill produces a very narrow particle size distribution which is not conducive to flowability of the cement or workability in concrete.  A minor addition of limestone or other fine filler (up to 5%) fills in the spaces between the evenly sized cement grains aiding fluidity and also taking up the space which would otherwise be occupied by water.  The effects of alkali and filler (LOI) are therefore likely to be different on the strength of concrete made to a given workability than on that of a mortar made to a constant water content.

As well as providing some indication of future performance of a cement, based on the various parameters measured, the true value is in prioritising the factors which adversely affect the quality.  A value can be put on the benefits of improving each aspect of the cement production so that sound financial case for improvement can be made.

Arthur Harrisson 2011