COV OF 28-DAY STRENGTH

Hi Norew,

To control cement strength variability, the main items to look at are;-

1. Sampling.
2. Testing.
3. Clinker quality (proportion of main clinker minerals and their reactivity,  free lime, LOI, alkali sulphate, use of AFRs and physical state eg. weathering, hardness).
4. Gypsum addition/quality.
5. Limestone or other mineral filler addition/quality.
6. Cement fineness/ particle size distribution.
7. Mill exit temperature (gypsum dehydration).
8. Grinding aid usage.

1. Sampling.

Samples should preferably be taken from the despatch silo(s) by automatic sampler. Manual sampling can introduce variability due to human error, or due to inadequacies in the manual sampling location or variability in the procedure.

2. Testing.

Strength testing variability should be minimised and measured by regular testing of a bulk cement reference sample. Preferably this bulk sample should also be tested by other cement laboratories as well. Of course such a bulk cement sample needs to be very well mixed and once homogenised it must be representatively split into 20kg samples kept in airtight buckets. The bulk sample only needs to be sufficient to last for 6-12 months, or whatever you determine is the acceptable lifetime of the sample based on long-term testing results ( ie acceptable range of testing parameter drift) Cement physical testing is often highly operator dependent, so an experienced, consistent operator is always preferred.

3. Clinker Quality.

Clinker quality variability (both chemical and physical) can significantly affect cement strength variability,  particularly the following parameters;-

(a)  Proportion of main clinker minerals, C3S, C2S, C3A and C4AF.

      The proportion of C3S has a strong effect on early strength (0-28d) while the proportion of C2S mostly affects late strength development (>28d).  C3A controls setting time but does not have a large influence on strengths. C4AF has little or no affect on strength. The variability of the quantities of these minerals should be minimised by controlling the variability of kiln feed chemistry (LSF, SR, AR), coal quality (ash content, specific energy and ash chemistry) and kiln operation (ie. free lime control).

(b)  Reactivity of the clinker minerals.

       Reactivity refers to the speed of hydration of the major clinker minerals, which can be influenced by;-

• degree of burning (mineral crystal size and reduced mineral reactivity due to over-burning.)
• presence of trace elements incorporated within the minerals' crystal matrix (eg, Fluoride, sulphur, alkalis)

       Reactivity variations can be minimised by controlling the amount of trace elements present in the raw materials/fuels and waste materials and controlling the burning zone temperature and burning zone residence time (kiln rotational speed and feed rate)

(c)  Free Lime.

      Free lime is a measure of burning degree, which in turn determines the quantity of clinker minerals produced and their crystal sizes and reactivity. So, control of clinker free lime between 1% and 2% will reduce the variability in cement performance caused by inconsistent burning..

(d)  Weathered Clinker.

      The use of weathered clinker (stored outside exposed to the weather) should be avoided because part of its clinker minerals will have become pre-hydrated over time, losing some of their reactivity. This will introduce significant variability in cement strengths unless costly and elaborate steps are taken to control the amount of weathered clinker used depending on its quality (LOI) and adjusting the milled cement fineness and/or limestone filler content accordingly.

4. Gypsum addition/quality.

Gypsum addition rate should be controlled by hourly measurement of the SO3 content of the cement. Cement SO3 should be controlled  to better than +/- 0.3%.  Gypsum quality should also be controlled because large, sudden changes in gypsum quality will increase SO3 variability, by the time it is noticed in the SO3 testing. and the problem rectified.

5. Limestone or other mineral filler addition/quality.

Limestone addition rate should be controlled by hourly measurement of the LOI content of the cement. Cement LOI should be controlled to better than  +/- 0.5%.  Limestone quality should also be controlled because large, sudden changes in CaCO3 content will increase LOI variability by the time it is noticed in the LOI testing. and the problem rectified. Variations in limestone content affect the clinker mineral content and hence contribute to cement strength variability.

6. Cement fineness/ particle size distribution.

Cement fineness affects not only strengths but also setting times and workability (eg bleed).  Controlling the cement fineness and particle size distribution (PSD)  will reduce cement strength variability significantly. Fineness can be determined by +45um (or +32um) sieve residue test or by surface area determination (or both). Sieve residue should be controlled to better than +/- 1% while surface area should be controlled to +/- 20 m2/kg.  Some plants also regularly measure their cement PSD and react to changes in this parameter (eg adjust ball charge, separator/mill internals maintenance etc)

7. Mill exit temperature (gypsum dehydration).

The heat generated in the cement mill will normally partly dehydrate the gypsum (CaSO4.2H2O) to hemihydrate (CaSO4.0.5H2O) and anhydrite (CaSO4) which are both more soluble than gypsum and therefore allow better control of cement setting time.  The degree of gypsum dehydration that occurs in the mill is dependant of the mill exit (gas) temperature (ideally 110-120 deg C)  Since fresh silo clinker is generally 90-110 deg C and heat is generated in the mill by friction, the mill exit temperature is usually controlled by using internal water sprays. However if cold clinker is used, the mill exit temperature may not be hot enough to suitably dehydrate the gypsum, resulting in 'flash set'. Flash set can also interfere with the hydration of C3S and C2S which will also increase cement strength variability.

8. Grinding aid usage.

Use of grinding aids may influence production rate, PSD and/or cement hydration properties (so called enhancing grind aids). Therefore if grinding aids are used their usage rate must be carefully measured and controlled to prevent introducing a variability in cement performance.

Regards,

Ted.

Dear Ted

nice presentation, thanks