Previous Post
Next Post




Kiln fuel is usually the largest single item of the operating cost of a cement works and, in a worldwide context, coal has always been the most commonly used fuel. This is because, on a cost per kilocalorie basis, it is usually less expensive than oil or gas. The large tonnages used by a cement works make it economically viable to install the necessary handling, grinding and firing equipment.

The term coal refers to a range of solid materials originating from vegetation There is a considerable variation in properties, arising both from differences in the source materials, and in their subsequent treatment. In general however, for a coal seam to exist certain conditions must have been satisfied. First, forest or other extensive plant growth must have taken place under swampy or water-logged conditions, to provide both for successive deposition of material and its biological degradation over a period of many years; and second, this must have been followed by superposition of silt or other inorganic material to provide both mechanical compression and a temperature rise.

The biological degradation taking place In the first stage will depend both on the types of plants and trees present, and the composition and oxygen content of the surrounding water. At the conclusion of the first stage however, (as a result of climatic or geological change) there will have been formed a layer of peaty material. The thickness of this layer will depend on the duration of the first stage, the so-called coalification process can only proceed thereafter by the application of mechanical load and increased temperature, over tens of millions of years. In general, the degree of coalificatlon will Increase with the magnitude of the loading and its duration.

The range of brown and black bituminous coals that we know today arises generally from the loads caused by various degrees of sedimentary deposition over decayed plant deposits of different geological ages. The degree of coalification is expressed as the rank of the coal, and changes in rank are characterised by certain chemical changes in the basic material. In particular the proportion of carbon and the calorific value tend to increase and the amount of volatile material (defined as the material driven off on heating to 900oC) tends to decrease as rank Increases. Broadly, rank of coal and age go together, but coals of the highest rank (e.g. anthracites) will have been achieved by geological loading and temperature increases greater than those associated with sedimentary activity一for example in mountain building movements, and/or igneous activity. Thus it is possible for very old deposits that exist near the surface to be of relatively low rank, through absence of the right kind of geological activity.


Estimates of total world deposits of coal (geological resources) and of that part of the resources that is recoverable (the reserves), are continually being revised in accordance with improvements in technology. For example Averitt (1960) estimated world resources at about 4.6 x 1012 tonnes (coal equivalent). In the mid 1970’s it was defined as 27.78 x 1012 Btu = 29.3 Gigajoule. Recovery of 50% of this would be equivalent to about 900 times the 1960 annual consumption.

What is technically recoverable depends, of course, on mining techniques which improve steadily. Economic criteria also currently favour greater recovery, by virtue of the high cost of other sources of energy, but Averitt’s figure of 50% with only about 1/16th of this recoverable. This estimate of reserves is nevertheless a substantial figure一about 250 times the world annual consumption for 1977 of 2.46 x 109 tonnes.


The range of compositions resulting from the various biological and geological processes was summarised by Dryden (1965) as shown in Table 1 (in order of increasing rank


Tvte of coalCarbonHydrogenOxygenMoistureVI
Brown coals and Lignites60-755.5-4.535-1750-30


Bituminous coals75-925.6-4.020-320-150-11



3.1 Major components

In this table, C H and O are expressed as a percentage by weight of the dry, mineral-free coal. Moisture content is of the coal as mined. Volatile matter (VM) is as determined by heating to 900”C, based on the weight of dry, mineral-free coal. Reactivity tends to decrease with increasing rank. A chart showing a general representation of the properties of British coals is given in Figure 1. Although the band does indicate the broad pattern of chemical composition, individual coals may deviate widely, particularly the dull coals (duraim). While calorific value is
commercially important in characterizing coal as a fuel, the other properties of volatile matter, moisture content (and mineral matter content) are often of greater practical importance. For example the low-volatile high rank coals, and in particular the anthracites, have high calorific value and low moisture content, but are of low reactivity and are not very suitable for kiln use because of ignition and flame stability problems


figure one Sey[ers Coal Chart 1951 All results on dry material matter free basis except capacity moisture and B.S. swelling numbe
figure one Seylers Coal Chart 1951 All results on dry material matter free basis except capacity moisture and B.S. swelling number
Towards the other end of the rank scale the coals have a high proportion of volatile matter and high reactivity, but unfortunately the greater similarity to the original vegetable matter is accompanied by increased characteristic water content. Much of this water is held in ultrafine pores in the coal structure. Anthracites and the higher rank bituminous coals have been found to have internal surface areas in the region of about 50 to lOOm2Ig, but more than twice these values may be found with the lower rank coals. The volume porosities of the black coals lie generally in the range 10 to 20%. These high values of Internal surface and volume, which arise from the biological origin and chemical structure of coals, lead to problems of water removal. Note that this microstructure is apart from the flaws and cracks associated with the layered structure, and from the geological forces that have acted on the coal.

Apart from volatile matter and water, a third factor having an important effect on coal combustion is Its mineral matter (ash) content. This can arise both from inherent plant minerals and from interspersed mineral bands, as well as fragments of seam roof and floor. After mining the coal may be taken to a washery, to separate out some of the minerals, but where an unwashed coal is acceptable there may be advantages in reduced water content and heat required for drying. Thus power stations may take coal of up to 35% mineral content although Increasing transport costs operate in favour of washing, the coal to a mineral, content of, say, 20%. BC! works In the UK generally use coals In the range of 4 to 20% mineral matter. The chemical composition of coal ashes is usually broadly similar to that of the argillaceous (clay or shale) components in the cement kiln reed raw mix, i.e. they are rich in the acidic oxides Si02, A1203 and Fe203 while比eir. CaO contents are usually low. Therefore, compared with gas or oil fuels, the kiln feed raw mix must be adjusted with, the proportion of clay or shale reduced, to maintain the required cement clinker composition. As a rough guide 2% ash content will reduce the LSF of the kiln feed material by about 1%.

Hence, the coal ash content should be maintained reasonably constant preferably to within about +1- 2%. High ash coals, up to 35 or 40% ash content, can be used in cement kilns provided that the ash content is controlled. For coals above about 15% ash content a stockpile-reclaim blending system is usually required.

Also, the coal must be ground sufficiently finely that the ash can be incorporated homogeneously into and react with the raw material feed in the kiln. For coals up to about 15% ash content a coal fineness of about 20% residue on 90 micron: is usually sufficient, but very high ash coals or low volatile coals need to be more finely ground. A second guide to fineness requirements, for flame characteristics, Is a 90 p.m residue equal to 50% of the volatile matter content. It can be seen that this would indicate lower residue requirements particularly for the high rank bituminous coals and the anthracitic coals.

3.2 Minor Components

We are referring here primarily to the volatile chemical constituents一chlorides, sulphur and alkalis. These constituents can have an important effect on the chemical composition of the clinker and the quality of the cement, and on the formation of rings or build-ups in the rotary kiln or the riser pipe of a suspension preheater or precalciner. It is the total input to the kiln of these constituents which matters, and the extent to which they can escape from the kiln system via the kiln exhaust gases and kiln dust. Therefore, the tolerable levels of these constituents in the kiln fuel is determined by:–

.  the amounts of them which are also entering the kiln via the raw

materials feed,

.  the type of kiln system,

Suspension preheater and precalciner kilns retain nearly all the alkalis and most of the sulphur within the kiln and the clinker product and the chloride forms a substantial cycle within the kiln system. Long dry and wet process kilns allow a significant proportion to escape and hence they are not so sensitive. The chemistry of kiln systems and cement clinker is a subject in itself and has been dealt with in another paper. It should be, noted, however, that suspension preheater (SP) and precalciner (PC) processes are very sensitive to chlorides, the maximum tolerable level of chloride input being about 0.03% Cl on clinker above which undue problems with KC1-bound deposits in the riser pipe will be encountered unless a hot gas by- pass is provided between the rotary kiln and preheater. The kiln feed raw materials commonly contain 0.01% Cl, so taking account of the fact that the raw material to clinker ratio is usually about 1.6 and for a SF or PC kiln the coal consumption is about 12t coal per lOOt clinker, there will be a tolerable maximum of 0.1% chloride in the coal fuel for many SP or PC kilns. Some coals can have chloride contents as high as 0.5% and thus be undesirable for SF or PC kilns.

Natural gas usually has a negligible sulphur content, so a change to coal or to heavy fuel oil fuel will increase the sulphur input to the kiln and the sulphate content of the clinker. Internationally traded boiler quality coals mostly have a sulphur content of 1% or less. Heavy fuel oils usually have a sulphur content of 3% or more. Taking account of the fact that on a kilocalorie input basis 1.5 tonne coal is equivalent to about 1 tonne of oil, a change from heavy fuel oil to coal fuel will usually reduce the sulphur input to the cement kiln system.


In view of the large range of properties encountered, a milling system set up to operate with a particular type of coal may not be able to accept a substantially different type一of perhaps identical calorific value. The factors that determine behaviour in a mill may be identified

.  by physical, chemical and mechanical properties


.     by grindability tests together with allowance for any special

circumstances such as water content or abrasivity.

The first method is used mainly for research purposes, but is also helpful as a guide to the practical problems lIkely to be encountered in coal milling.

4.1       Coal Storage

The quantity of coal to be stored depends on the works consumption and the method of supply. For instance, a plant located close to a coalmine would normally require less on-site coal storage than a works which is supplied by ship loads of say 15,000- 20,000t. If the coal is of a reasonably consistent quality and less than 10 or 15: ash content, then a simple heap at the works with a front-end loader to feed it to a 12 hour capacity coal mill feed hopper should he adequate. If the coal is of a higher and variable ash content, or if a number of very different coals areto be used, then a layered blending stockpile with bridge or side-scraper type reclaim will be required. This type of reclaim is less expensive on operating costs than a front-end loader.

Whether or not the stockpile should be covered depends on the climate and the environmental requirements. The heavier the rainfall the more worthwhile itis to cover the coal stock, also if the climate is dry and windy a cover will avoid the nuisance and loss of coal by windblown dust.

4.2 Variation of Physical and Mechanical Properties with Rank

The way in which chemical properties vary with rank has been indicated briefly in section 3. In general these properties are quoted with reference to the dry, pure, coal (i.e.,. separated from mineral matter). Such a complete separation is not possible, or desirable, however physical or mechanical tests, do impose various restrictions. For example, the coal is usually tested In the ‘air-dried’ condition, having been stabilised in a laboratory atmosphere. Tests of compressive strength, f or example, must take account of orientation, because coal is generally stronger in a direction perpendicular to the main plane of the seam一the bedding plane. Measurements of such properties as strength and elasticity are usually made on specimens which have been prepared from the lump by means of a diamond or silicon carbide slitting wheel.

Compressive strength values have been obtained in this way world-wide for many coals, initially because of the practical importance of this measurement in mines where coal pillars were left to support the roof, but subsequently in the development of theories of the strength and breakage of coal. Figure 2 shows a typical set of values of both compressive strength and an impact strength index for British coals as functions of VM content (Drown&Hiorns 1963). The parabolic relationship is typical and arises from the variation of Volatile Matter with rank, so that strengths are effectively plotted against rank. The high rank coals (low VM) tend to be hard, strong and brittle.

The low rank coals show increasing toughness as volatile matter Increases, due to the woody nature of the partially transformed material.

The strongest coals (the anthracites) are about equivalent in compressive strength to a moderately strong limestone. Strength is very much less in tension, a typical value being around 1.5 N/mmz for a hard coal. Mill wear rate is also related to compressive strength, as will be demonstrated later.
Other mechanical and physical properti~ that may be expected to affect the performance of a mill with a specific coal are less easy to find. Little information is available on hardness, but some measurement on British coals by Parish (1970) suggest a micro-indentation hardness range from about 450 N/mm* for anthracites to below 200 N/mm* for medium rank coals – equivalent to say from a rather hard calcite down to gypsum or rock salt. Coefficients of friction, likely tq affect the performance of a ball mill for example through the lift of the charge, and in other ways, have been measured by Pomeroy (1964) and the values quoted are high, lying between about 0.4 and 0.6 but are sensitive to operating conditions: much lower values may apply with some minerals.

Overall, the general picture is of a range of materials going from quite soft to moderately hard, from fairly tough and woody to brittle; and with the fragile coals in the middle of the rank range being heavily pre-cracked. In contrast the highest rank coals tend to have a close, compact structure with however an extensive distribution of ultrrafine pores.

4.3 Effects of Mineral Matter

The range of minerals found in association with coal deposits is extremely wide (Cooling 1965) and in consequence, the range of properties is substantially greater than that of coals. In particular, whereas the hardness range of coals is about 200 to 450 N/mm2, it extends to 1400 N/mm* and occasionally above for the associated minerals, of which the hardest commonly found are quartz and pyrite. Since the wear rate of mills is largely determined by the hardness of the feed (Parish 1963), the relative, contribution of the coal substance to wear is often small if, for example, the feed contains appreciable quantities of quartz or pyrite.

FiGURE 2 Strength and Rank
FiGURE 2 Strength and Rank

Figure 3 shows this effect for mineral particles smaller than 75 um, appreciably greater wear rates will be obtained with larger mineral particles. If facilities are available it is always worthwhile to have the free quartz and pyrite content determined, both as to quantity and particle size as part of the coal assessment. Alternatively an abrasivity test may be carried out (see section 5.3)

4.4 Coal as Received Fineness and Free Moisture Content

Free moisture content is that part of the total coal moisture content which is not inherently bound or absorbed into the coal substance, but which is present as free water between the coal particles. If the coal has a high proportion of fines and a high free moisture content, it will tend to stick onto conveyor belts and inside elevator buckets and to stick and not discharge regularly from hoppers and feeders. This can cause irregularities in the operation of the coal mill and sometimes the kiln. Hoppers need to be properly designed to suit the flow properties of the reals which will be employed -steep cone angles, avoidance of corners, generously sized coal discharge aperture, and smooth internal surf aces. Coals with a < lmm size fraction of 30% or more can prove extremely difficult to handle through a poorly
designed system. This is made worse if free moisture content exceeds 7-8%.

A total coal moisture content of more than about 15% can cause a significant increase in the fuel consumption of an SP or PC kiln if a direct firing system is employed.

figure 3 Effect of Mineral Concentration on Abrasive Wear by Quartz, Siderite and Calcite Particles (minus 75pm)
figure 3 Effect of Mineral Concentration on Abrasive Wear by Quartz, Siderite and Calcite Particles (minus 75pm)


5.1     Grindability Tests一the Hardgrove Test

In the grindability test a sample of the coal is ground in a model mill一usually a batch mill一and the product assessed in some way for degree of size reduction. Of several tests in existence, the most common, and the only one considered here, is the Hardgrove Test (ASTM 1971).

This test Is also adopted by the ISO. It was Invented In about 1934 by R M Hardgrove and uses a small ring-ball mill shown in section in Figure 4.

The test requires the coal sample to be representative of the original consignment, and to be prepared according to certain standard conditions. The object of the preparation is to obtain as much as possible of the original coal sample in the size range 1180 to 600 1.I.m. This is the size used for the test grind, much care is needed in the preparation because the feed will be a mixture of coal components and minerals. All these constituents will have different resistance to breakage, and no matter how carefully the sample is prepared, there will always be a tendency for the weaker components to break down more readily. Thus If the original sample is sized 25mm down, and it is all broken down to a top size of 1180 1am with the object of extracting the 1180 to 600 pm fractIon (14 to 25 mesh BS), it will be found that the weaker, more readily ground components will tend to concentrate in the finer parts of the crushed sample一so the 1180 to 600 pm fraOtion will contain a greater proportion of the stronger components, and will not be representative of the original sample. It has been found that this effect can be minimised by using a number of stages of gentle breakage to reduce the original sample to the maximum size of 1180 pm. Even so, it often happens that only 50% of the original sample can be recovered in the desired size range, and there may be appreciable differences in composition between the original coal and the sample used for the grindablilty test.

The test requires 50g of the coal to be ground for a specified period In a small laboratory mill under standard conditions, and the product is then sieved on a 75 p.m sieve (full details are given in the Standard). If P Is the percentage of the product from the test that passes the sieve, then Hardgrove Index== 13+3.465 P.

The formula has no theoretical significance, but was originated by Hardgrove to give an index of 100 to a standard American coal一which is now no longer available. It would be more sensible to quote simply the percentage P passing the sieve, as the grindability index, but although several attempts have been made to popularise this, the Hardgrove scale has remained in favour一probably because it is a convenient whole number scale which runs from values of about 25一30 for the very hardest coals, to 120一130 for the softest.

The relation between Hardgrove grindability and coal rank is of only a broad Kind, but shows the characteristic parabolic form. Figure 5 shows volatile matter (dmf) and Hardgrove index for a number of representative coals. The spread of points is typical, and it is obvious that grindability cannot be forecast from volatile matter. For example for a volatile matter content of 32% Hardgrove index could lie anywhere between about 60 and 95.

Similar considerations apply to other coal parameters as well as volatile matter. It can be noted here that for a given mill the relationship between Hardgrove index and output can be useful in predicting the effects of changes in coal specification.

5.2 Criticism of the Hardgrove Test

The defects of the Hardgrove test arise as much from our own lack of understanding of the grinding process as from the test or machine itself. For it to be of use to the plant engineer, in relation to a specific mill, he must plot mill output under standard operating conditions against Hardgrove Index for a range of coals. In this way a calibration curve is constructed which can eventually be used to forecast mill output with a new coal, from a measurement of the Index. However, there are several defects of the test, two of which may be noted here as examples:

Example 1. Batch versus continuous milling:

In a continuous mill, particularly with air-sweeping, an equilibrium system is luilt up due to internal separation of particle sizes and preferential grinding of softer materials, so that the mill contents are often not of the same composition as the feed. The Hardgrove test does not allow for this possibility.

Example 2. Effect of water:

The Hardgrove test uses a sized coal air-dried under laboratory conditions, and this is intended to approximate in proportion to coal dried in the full-scale mill by a hot gas stream. In practice the heat available to dry the coal is often inadequate, and depending on the type of mill, and the amount of water in the feed, mill output may be seriously affected through inadequate drying. The effect of incomplete drying is to alter the transport properties of the particles in the mill, a properly dried coal (I.e. one with not more than one or two percent over inherent water content) should flow freely, the flow properties are quite different If water Is present. As an example of this Figure 6 shows the behaviour of a small open-circuit gravity fed laboratory mill, working at constant feed rate but with a series of feeds of increasing water content. Compared with a dry feed, addition of a relatively small quantity of water substantially reduced the extent of breakage by changing the pattern of flow of material under the grinding elements. Only by increasing the water content to the level at which the water acted as a transporting medium was the mill operation restored almost to its effectiveness with dry feed.

The type of curve shown in Figure 6 has been obtained with other laboratory mills, but the minimum occurred at different moisture contents一for example, at around 16%, for a ball mill. Experiments with wetted coal in hoppers gave curves of flow against moisture content similar in form to Figure 6, but the minimum occurred at around 10%. It is well known that the transport properties of wet materials depend on their particle size distributions, and the fines content will determine the surface area and have a strong influence on the effect of water.

Thus the location of minimum flow properties in the ball mill at 16%; may have been a consequence of the high proportion of fines in this type of mill. Conversely the experimental mill of Figure 6 was designed to prevent the hold up of fines in the grinding zone. Certainly inadequate drying of coal in a ball mill Installation will reduce output and may lead to blockage of the mill. However, it has been shown in many installations that if the exit gas temperature of the mill Is maintained at 70-750C, satIsfactory operation should generally be obtained.

There are many other examples of the importance of transport effects In mills. In the late 1930’s, in-the-mill coal drying was relatively new, and Schulte (1939) quoted Improvements in tube mill performance as a result of mill drying and air screening: reduction in mill length from about 11 metres to about 5 metres, and reduction in mill energy consumption of about 30%. Conversely, there are numerous published examples of difficulties experienced due to Inadequate drying, or to materials that tend to cake or agglomerate within the mill, or coat the grinding elements. All these can be regarded as aspects of the problem of

transporting particles through the mill一and with coal grinding this most frequently             ・
arises from the presence of water.

Another factor that can be of considerable importance is mill condition; with some designs, a worn mill can be substantially less effective than a new one. Some ways of reducing mill wear are considered later.

It will be seen from these examples that a coal grindability test cannot be a precise guide either to the energy needed to grind a particular material, or to the capacity of a given mill with that material. In any case, the grindability test is only a comparative measure of ease of grinding, and is of no use unless a curve relating mill capacity to grindability Index Is already available. General relationships such as that shown in Figure 7, which has been Plotted from data given by Spiers (1962), are available, but are of little practical use. A reliable curve for a given installation can only be obtained by full-scale experience, and every opportunity should be taken to update the curve as new feeds become available.

5.3 Abrasion Tests

So far as operating costs are concerned, the abrasion test may be as important as the grindability test. Even greater care is needed in sampling and sample preparation because, as noted above, the principal contributions to mill wear are likely to come from hard minerals, which may be present only in small quantities. These materials may also be irregularly distributed. As with the section on grindability testing, only one test will be referred to here, but it is probably the most widely used. This is the test originated by Yancey, Geer and Price (1951)

which employs apparatus similar to that shown in section in Figure 8. The coal is first reduced to a top size of 6 mm(0.25 in.) and a 4 kg. sample placed in the steel pot. The shaft holding the four arms, on which are mounted four removable soft iron paddles, is rotated for 12,000 rev, at 1500 rev/min. The paddles are weighed before and after the experiment, and the total weight loss in milligrams is the abrasion index.

The test Is relatively easy to carry out, and has been shown to correlate quite well with measurements of hard metal loss in a laboratory roll crusher, using a radio tracer method (ParIsh 1970). It was used extensively both by the original authors in the United States and also in a survey of South African coals. The version of the equipment used at the Greenhithe Technical centre differs in a few respects from the original;

In particular

.           an improved bottom bearing dust seal,

.           paddles constructed of Swedish iron annealed to a precise specification and

.           a more powerful motor.

The equipment is used to measure the abrasivity of clinkers as well as coals; typically, weight losses are in the range 13 to 175 mg for coals, and around 350 to 900 mg for clinkers.

It may be noted at this point that in coal grinding, wear may not be entirely due to abrasion. Under conditions of inadequate drying it is possible for some coal minerals to exert a significant corrosive action on the grinding elements, and a combination of corrosion and abrasion will produce significantly greater removal of material than either effect separately. This emphasises once again the importance of drying in coal grinding.


In the UK medium speed ring roll and ring ball mills are used for the whole range of coals and are now used on all works having superseded the older ball mills, partly because of greatly superior drying ability, quicker response to an Input change, greater safety and lower power.

From sections 1-5 it will be seen that in coal grinding it is unlikely that major problems will be encountered in achieving the required size reduction (to about, say, 10 to 25%+90um) provided certain rules are observed. Summarising, the types of likely problem may be classified as:

6.1 Types Of Coal (As Expressed bv Rank or Grindability)~

High rank coals tend to be hard, relatively difficult to grind, to be unreactive and therefore possibly giving rise to ignition and flame stability difficult. As indicated earlier finer grinding may be necessary if volatile contents are low. If a high-speed mill is to be used, wear problems will also be encountered, possibly
severe. Drying is not likely to be difficult. Conversely low rank coals (brown coals, lignites) are reactive, have high water contents, and may be difficult to dry adequately. They may be difficult to grind through halving a fibrous nature and also by inadequate drying. Wear due to the coal substance should not be a problem.
Medium rank coals are of course intermediate in properties except for grindability,
which is generally easier, and the possible presence of a greater proportion of fine
material holding surface water and giving problems of drying and transport.

6.2 TYpes of Associated Minerals:

The principal problem here is mill wear, arising mainly from the presence of more than a few percent of hard and/or corrosive minerals. Some minerals may also give rise to problems associated with coating of the grinding elements. For the latter, impact mills may be more suitable than slower types; for the former, if the problem is severe, it may be necessary to reduce the mineral content of the coal, or choose another source. In any case a slow or medium speed mill will be necessary, in some cases for particularly abrasive coals a ball mill may have to be considered in
preference to increases with fineness of grind

6.3 safety

Safety problems in coal handling and grinding installations arise from
– the ability of the coal surface to oxidise in air and to produce heat in so doing
– the fact that suspensions of fine coal dust, (except for high rank coals) will explode in air given an adequate ignition source and suitable concentration.

6.3.1     Self Heating

The tendency of coal dumps to heat up is well known: according to Hall (1979) It was first observed in 1686, and the number of subsequent publications is considerable. It is well known that, since the temperature rise is due to oxidation, it is necessary to minimise oxygen access by building the heap in relatively thin (50 cm) layers, followed by compaction. The rate of oxidation will depend also on coal rank and the surface available一I.e. the finer the coal, the faster it will oxidise and the greater the heat produced. It can be seen from the above comments that a low rank coal will be more at risk of stockpile fires, and, conversely, bituminous high rank coals should be less reactive. Some care in managing open stockpiles is normally all that is required.

The high surface area of finely ground products renders the self-heating problem more serious in coal grinding installations, danger spots being hoppers horizontal or slightly inclined pipes, and filter bags一In fact, any points at which the ground product can accumulate on shutdown or during operation. All hoppers used for pulverised coal should be fitted with temperature sensors and means for flooding with CO2 or water, and on shutting the mill down the gas flow should be maintained to clear the system so far as is practicable.: Conveying pipe gas velocities must be designed to ensure that coal deposition does not occur In horizontal sections of pipe, typical velocities being around 30 rn/see, with a minimum of at least 24 m/sec. Although the high rank coals are also less reactive in the finely ground state the above design considerations should be applied rigorously in any Installation handling fine coal.

6.3.2 Explosions

The need for avoidance of spontaneous heating in milling systems arises from the possibility that a hot deposit can act as an ignition source for a coal dust cloud during start-up or shut down, these usually being the only times at which dangerous dust concentrations in the explosive range will be encountered. Experience has shown that air/coal ratios less than 5kg air/kg coal can be considered safe. Correspondingly, deposits of coal dust around the plant, which may be dispersed into the air by a primary explosion and so cause secondary explosions, must be avoided. It is assumed here that in any reputable plant design, adequate provision of explosion relief vents will have been made, and it is only necessary for the operator to ensure that these are not obstructed in any way.


The predominant firing system in the UK is still direct firing with the mill output passing “directly” to the kiln, as shown at the top of Figure 9. The reasons for this concentration on one type of firing system were that it was safe, and cheaper than either of the alternatives shown in Figure 9, both of which involved storage of the ground product in a hopper. These示tems either use the mill sweeping gas to take the coal from the hopper (semi-indirect firing), or take the de-dusted mill gas to atmosphere, and use a, separate primary air supply (fully indirect firing).

Both forms of indirect firing have been used, for reasons of better control of coal throughput and (for fully indirect firing) the independence of the primary air supply from the demands made by the mill. The systems were mOre expensive both in capital cost and to install and operate when compared with direct firing, and hopper fires were not unknown.

With improved safety devices, and the increased use of low oxygen preheater gasses for drying, fully Indirect firing has become the preferred system for new installations. This is due to the advantages in kiln performance, that is, reduced fuel consumption, Increased kiln output and Improved cement quality. The case for conversion to indirect firing from existing direct fired systems is not as clear cut. The costs can prove difficult to justify unless the conversion offers clear possibilities of using a cheaper lower grade fuei or possibly, in the future, a means of complying with environmental requirements.

The main features of the various firing systems are summarised Table 2 there are other variants of the semi indirect system which are not covered within the context of this paper.




Three types of mill are commonly used for coal grinding in the cement industry
these are:-
●Vertical spindle mills
●Ball mills
●High speed (attritor or impact type) mills

8.1     Vertical Spindle Mills

8.1.1 Mill types

Vertical spindle mills occur as two types, ring roller or ring ball mills. A typical ring ball mills are shown schematically in Figures 10 and 11,and figures 12 and 13 show variations on the ring roller type. The mill requires a hot gas supply for drying. Raw coal is fed centrally to the rotating grinding table, the coal is ground as it passes under the grinding balls or rollers and is then picked up in the rising hot gas stream. Drying takes place as the coal is lifted inside the mill. The gas then passes through the vanes of a separator, the vanes give the stream a tangential motion and coarse particles are separated and returned to the grinding table. Modern coal mills of both vertical spindle types are now supplied with a dynamic separator with a high efficiency cage type rotor, figure 14 shows one of these modern designs and figure 15 shows the conversion of an earlier ring ball design with a dynamic separator. These mills have the advantage of being able to operate at the very low residues sometimes required for low volatile fuels. A larger separator can be used to enhance the drying capacity.

Fine coal leaves the mill with the gas stream, the coal fineness is adjusted by altering the angle of the separator vanes, or the rotor speed. Coal drying is ensured by controlling the coal mill inlet temperature, to give a target mill exit temperature, usually between 70 and 900C. Drying capability varies according to mill circuit design, Table 3 summarises the position.


Direct firing一low primary air (<20% PA) suction mill.0-10% moisture in coal
Direct firing一low primary air (<20% PA) pressurised mill0-15% moisture in coal
Indirect firing一standard separator.probably 0-18% moisture in coal
Indirect firing一larger separator0-25% moisture in coal



This type of mill is designed typically for a gas flow up to 2.5 kg of air per kg of coal, at maximum output. Higher figures (up to 3.5 kg/kg) can occur in direct/semi indirect firing particularly if the mill is run at lower capacity.






Mills for indirect firing systems can be run at higher volumes if drying requirements are high. Where drying requirements are low, they can operate down to 1.7-1.8 kg/kg.

There are advantages in minimising the mill airflow, thus reducing the primary air level. However mill output and drying capacity can be limited, possibly affecting kiln output.

Drying capacity at a given gas flow is usually limited by the fire risks of higher mill inlet temperatures. A provisional maximum of 300oC should be considered however this figure will be dependent on coal volatility, and the beahviour of each mill type with regard to coal falling through the port ring (a common seat of mill fires). Mills operating on kiln gasses (ie low oxygen levels) are less sensitive to high inlet temperatures.

If lower air quantities are used the fan pressure capacity must be adequate to ensure sufficient momentum at the burner nozzle. Velocity through the mill port or “throat” ring must be sufficient to avoid coal building up in the gas path below the grinding table. This can be a cause of fires in the coal mill. Calculated velocities at the throat vary between 50 and 80 rn/sec. Figures towards the top of this range can represent a waste of available f an power. At the lower end the risk of build up and fires increases.

The load on the grinding table in a ring roller mill Is provided by the weight of the grinding balls and by a loading ring which presses down on all the balls. In a ring roller mill the individual rollers are tensioned on to the rotating table.

The grinding force in both mill types is adjustable using pre-tensioned springs or hydraulic rams.

8.2      Ball Mills

Ball mills used for coal grinding have most of the features of conventional ball milling systems. They used to be more common but the duty has to a large extent been superseded by modern vertical spindle mills. Ball mills have been used in both direct and indirect systems. The most common designs using fully air swept mills with drying chambers. As with raw milling drying capacity can be a limiting factor and some external predrying of coal may be neccessary for moistures above about 12%. The system has a slow response to changes in feed rate so the control of fuel flow in direct firing systems can be difficult. The large airflows required for air sweeping can result in high primary air quantities and explosion risks in non-inert (ie atmospheric air) systems. The more complex system requirements usually with a closed circuit mill, static separator, collection cyclones etc increases the risk of static coal build up and fires (figure 16). There is also a significantly higher Initial capital cost and some increases in operating cost.

The ball mill becomes a potentially attractive alternative in situations where the coal source is highly abrasive. Under these conditions the frequent replacement of vertical spindle mill internals can tip the economic argument in favour of the ball mill. The major issue is the effect the longer vertical spindle mill stoppages can have on kiln run time. Ball mill grinding charge can be replaced/topped up on a short stoppage and liner repairs can usually be phased to coincide with major repairs once or twice annually.

Closed circuit ball mills operated with an indirect firing system have been installed recently for highly abrasive coals. Developments in vertical spindle mill technology will continue to make this a less common occurrence.


8.3 High Speed Mills

8.3.1 Atritor Mills
The earliest type of high speed mill used in Blue Circle was the Alfred Herbert atritor mill (Figure 17). This device combined the drying grinding and firing fan operation in a single unit. The assembly shown in Figure 17
comprises a single shaft on which are mounted:-
-Precrushing zone with swing hammers
-Grinding zone with fixed impact blades
-Grinding zone with fixed and moving pegs
-Rejector plates a crude separator to return coarse particles to the grinding zone

-Fan assembly

The use of this machine for coal grinding has been largely discontinued in favour of the vertical spindle mill its specific disadvantages were, high wear rates, high primary air, mechanical unreliability (due to tramp metal), and poor control on product fineness.

A development on the atritor principal is the Saxifrage mill (Figure 18) this design attempts to retain the simplicity of the atritor while eliminating some of the the problem areas. The grinding is effected by a single set of swing hammers reducing the risk of tramp metal damage. Classification is carried out in the mill using a separator rotor which is driven faster than the main shaft through a jackshaft and pulley arrangement giving better control of product fineness. The mill operates with an airflow generated by an external fan. This could be a clean gas fan after a bag filter, or a dirty fan (firing fan). The design is not at present considered suitable for operation as a pressurised system.

Although the wear rate and high primary air issues may still apply the assembly could have potential applications in areas where a small grinding unit is required. This could be for auxiliary firing in a preheater kiln or possibly for substitute solid fuel injection in precalciner or preheater kilns.

8.3.2 The TAS Mill

The TAS mill which has been in use in two BC! preheater kiln plants for auxiliary firing since the late 1980’s is shown in figure 19. The design uses a vertical rotor assembly with each rotor disc operating as a grinding stage. Wear blocks on each rotor disk are replaceable, the grinding process is said to take place by an attritive action In the turbulent zone where the coal Is thrown off each disk.

There is no specific “external” residue control mechanism although airflow and coal throughput will contribute to the final product fineness. Longer term wear allows the internal clearances to increase and residues increase as a result. Clearances, and residue performance, can be corrected by adjusting the rotor disks on the centre shaft, however, experience has shown this to be a difficult operation and as a result Infrequently undertaken.

The mill has an integral fan although for any significant transport distance additional fan capacity is required. The mill has high wear rates and requires significant protection from tramp metal. Overall experience has shown these mills to be very maintenance intensive and future installation for比is type of duty would require careful consideration of any possible alternatives before the TAS.

9 sampling

Mill products should be sampled at regular intervals for determination of size distribution (normally the residue on the 90 pm and 300 pm sieves) and moisture content. Samples are normally obtained by inserting a probe into the firing pipe; because of stratification of sizes and gas velocities within the pipe a single-point sample obtained in this way is unlikely to be representative, and it is preferable to take samples at several points by traversing the pipe, although even this method is likely to give a biased sample. However, provided that the procedure is carried out in a consistent manner, the results should be sufficient to show any deviation from normal performance. The sample point should be situated in a straight length of pipe, and, as far as possible from any bends, and from the fan if this is after the mill. The traverse should be in the same plane as the nearest upstream bend.


Hiorns F.J.”Coal Preparation’,一Blue Circle Cement Technology Course Dover P.L. “Coal Preparation’,一Blue Circle Technical Training Paper

American Society for Testing and Materials., 1971. Grindabillty of Coal by the Hardgrove Machine Method D 409-7 1

Brown, R.L.&Hiorns, F.J. 1963. ‘Mechanical Properties’ in Lowry, H.H., ed ‘Chemistry of Coal Utilisation一Supplementary Volume’ Ch.3. pp 119-147. John Wiley&Sons, New York.

Cooling, D.R., 1965. ‘The Hardness of Coal and Its Associated Minerals’:BCURA Mon. Bull 29. 409-435

Dryden, 1.G.C., 1965. Kirk-Othmer Encyclopedia of Chemical Technology 5, 606- 678. John Wiley&Sons and Interscience Publishers, New York.

Hall, D.A., 1979. ‘The Storage of Coal’;Coil .Guard 227 (6), 331-40

Parish, B.M., 1970. ‘Wear of Puiverisers’ BCURA Research Report 363: ll3pp+ 55 /Figs. NCB Publications, Hobart House, Grosvenor Gardens, London SW!

Pomeroy, C.D., 1964. ‘Friction between Coal and Metal Surfaces’ Coil. Engng. 4! (2),67-72

Schulte, F., 1939. Die Warme 62 (52)

Yancey, H.F. Gear, M. R.&Price, J .D., 1951.The Abrasiveness of Coal and its Associated Impurities’, Min. Engng. 3, 262

please contact us if you have any inquiry


Contact US

Previous Post
Next Post

Leave a Reply