Ball Mills Training Manual: Dirección Técnica
Ball Mills Training Manual: Dirección Técnica
Ball Mills Training Manual: Dirección Técnica
INTRODUCTION
In order to diminish wear, some mills are being equipped with a rubber lining
in the second compartment.
Since the cost of energy used in cement production has risen, alternatives
have been sought to lower production costs.
At present, the new production lines have dry-process kilns and closed circuit
mills when grinding cement with a fineness over 3,200-3,300 Blaine. In special
cases, fast-hardening cements can be produced with up to 5,000 Blaine. In raw
meals the requirements are 85-95% through 90 microns. Open circuits are scarcely
used in cement grinding, except when the fineness requirements range from 2,600
to 3,200 Blaine.
Fresh End
Feed Product
Mill
OPEN CIRCUIT
Separator End
Rejects Product
Fresh
Feed
Mill
CLOSED CIRCUIT
Figure 6.9 Open circuit and closed circuit grinding systems.
same retention in the control mesh, but the product will have a lower retention at 145
microns.
Single chamber mills are used little in cement grinding. The Barrientos plant
has a KVS single-chamber mill in operation for cement grinding.
The advantages of two-chamber mills include dividing the ball sizes into two
different chambers, little restriction to the flow of gases, and water can be injected
into either of the two chambers. It is used for raw meal grinding when the feed
granulometry is very coarse or inconsistent, the diaphragm does not allow the larger
pieces to pass into the fine grinding section. This type of mill is the most used for
cement grinding because of the advantages it offers.
This type of mill is generally used with an open system producing very fine
cements. Figure 6.12 shows a three-chamber mill. They are not used for raw meal
grinding because of the strong pressure drop.
In cement grinding, the fresh material is introduced into the coarse grinding
compartment. For raw material grinding, the material is fed directly into a drying
chamber before passing into the coarse grinding compartment. After this grinding,
the material is discharged through holes found in the shell periphery and is fed into
the separator by means of an airslide and an elevator. The particles rejected by the
separator are fed into the fine grinding chamber and the fine material is captured by
cyclones or dust collectors. The material discharged from the fine grinding chamber
is also fed into the separator. The ball loads in each compartment are the same as
those in a mill with end discharge. Figure 6.13 shows a central discharge mill.
This type of mills are used to grind raw meals with low moisture percentages
and the majority of the cements, taking advantage of the material’s slow passage.
Material discharge in these mills is carried out solely be mechanical means. The
ground material flows because of the pressure exerted by the fresh feed. The final
end of the mill is equipped with plates that let the material through, holding back only
the grinding bodies. Additionally, the majority have plates radially located to help the
material flow.
A given amount of air is forced in this type of mills for several purposes:
x Dedusting the mill inside, thus removing the fine material.
x Cooling the material and helping to remove the steam generated by the
cooling water or raw material evaporation.
The normal air flow operation speed within the mil are around 0.8 to 1.2
m/sec, handling 10% of the material in the air flow.
x Separator discharge.
In the majority of the mills, the material that is rejected by the dynamic
separator is introduced into the first mill chamber along with the fresh feed. When
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 10/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
using a central discharge mill, it is possible to send this product to either the first of
the second chamber. Under normal operation conditions, 10% of the rejected
product is sent to the first chamber. In case the feed has a high humidity
percentage, a larger portion of the material is conveyed to the first chamber to
eliminate as much humidity as possible. This situation is shown in figure 6.16.
Figure 6.16 Grinding system with central discharge mill showing the possibilities to
handle separator rejects of material.
By letting cold air into the separator, the final cement temperature can be
controlled and kept between 80°C and 85°C. The disadvantage this arrangement
poses is a complex control of air flow because it is necessary to maintain a constant
air mass through the separator so as not to alter the quality of the product. The
system is shown in Figure 6.17.
Figure 6.17 Grinding system with gas recirculation to separator and independent
mill venting.
Figure 6.18 Grinding system with gas recirculation to separator and a single
collector.
This type of arrangement is adequate when wet additives are used and the
material temperature is not very constant. The disadvantage to this arrangement is
the need to use a cement cooler when working with the maximum return
temperature.
x Mill collector and air-sweeping fan.
In the last two arrangements, the mill is dedusted directly into a collector.
Another alternative to this array could be venting the mill through the separator, as
shown in figure 6.20. Following is a comparison of the two mill venting alternatives.
This comparison is interesting when the product’s minimum temperature is
controlled so as to avoid using the cement cooler. Figure 6.20 shows a ventilation
system through the separator, and figure 6.21 shows a setup where ventilation is
carried out directly to a collector.
To compare both systems, specific consumptions have been calculated,
based on typical parameters of pressure, temperatures and gas flows. Figure 6.19
shows the ratio of system 6.20/6.21 en fan consumption, the gas flow through the
separator and the total volume of gases to dedust as a function of the temperature of
the cement in the separator-ventilation system, maintaining at 75oC the final
temperature of the cement in the collector-ventilation arrangement.
1.7
1.6
1.5
1.4
1.3
Total air volum e
1.2
Separator air flow
1.1
Energy consum ption
1
0.9
0.8
0.7
0.6
75 77 79 81 83 85 87 89
°C
Temperature of fine part. and rejects
Figure 6.19 shows that, in the case of mill venting through the separator, by
keeping cement at a temperature of 70oC, the fan’s energy consumption is 70%
greater than the consumption of the two fans used in the independent mill venting
arrangement. At the same cement temperature (75oC), a larger volume separator is
needed because of the 30% increase in air handling in the latter setup. The total
dedusting volume is practically the same in both arrangements. The greater need
for air is because the amount of air handled through the mill that passes through the
separator does not contribute to cement cooling.
Figure 6.20 Grinding system with mill venting through the separator.
Consequently, it is necessary to increase the demand for cold air to cool the
material discharged from the mill. The cold air must be introduced into the separator
by means of a gate which generate a pressure drop equivalent to the mill’s pressure
drop. This fall in pressure, and to a lesser degree, the greater volume of air being
handled through the separator are the cause for the higher energy consumption that
characterizes this arrangement.
Figure 6.21 Cement grinding with independent draft through the mill and through
the separator.
It can also be concluded that for the same size of separator used in both
arrangements (that is, the same air flow through the separator), the separator-
ventilation setup operates at temperature 5ºC to 10ºC higher in the rejects and in
cement that the alternate arrangement.
Satellite
Kiln
The hot air needed for drying the coal in the mill can be taken from the
residual kiln gases, the clinker cooler, or from a hot gas generator. The fan suctions
gas from the mill and discharges it along with the pulverized coal, serving as primary
air to the burner. The disadvantage of this system is that false air is injected into the
kiln, resulting in high heat consumption. Additionally, from a process point of view,
mill production should vary according to the kiln’s fuel requirements. There is also a
great interdependence between kiln and mill production. Any disturbance in the
operation of the mill can result in kiln shutdown as it has no coal reserve.
This is a variation of the direct system. In this one, coal is ground with an air-
swept mill. The pulverized coal is separated in a classifier and the fine particles are
picked up by a cyclone. The system fan operates with clean gas, part of which is
sent to the primary air fan. Coal supplying is a more independent from mill
production. The separator cyclone supplies pulverized coal to the burner, and acts
as a damper for any production variation, thereby preventing any upset to the kiln.
Satellite
Kiln
Satellite
Kiln
The advantages of an indirect injection system are: mill and kiln production
are independent, its possible to feed various kiln systems at a time, primary air and
coal injection can be optimally regulated.
The disadvantages of these systems are: greater initial investment on
equipment, the need for a high performance filter, and inert gas equipment for fire
safety.
The mill ends are a solid cast iron piece. These ends are made of the conical
part and the axle joints (should it have this type of support) and the shell joints. The
shell joints can be either flanged or welded. The conical part of the mill ends made
nowadays are completely smooth; the reinforcement nerves, used in the past, cause
considerable harm to radial tension resistance. The front sides of the mills are
submitted mainly to bending forces; traction and compression forces are minimal.
6.4.2.1 AXLE/BEARING
In general, ball mills in the cement industry have smooth axle/bearing
supports. They consist of two smooth surfaces, one of them moveable (axle, moves
as the mill turns) and the other is stationary (bearing, to the foundation), made of
wear-resistant material covered with oil. Their functions: hold the mill; enable its
rotation and form a lubricating film that can cover the mill; and serve as an oil-lift
device. One support, on the transmission side, is stationary and acts as a guide
support, whereas the other has tolerance to axial movement to allow equipment
(mill) dilation. Figure 6.25 shows a diagram of the support.
Over the years, different grinding systems have been developed with a
sustained tendency to improve the process, to make it more efficient. Selecting a
grinding system basically depends on the required quality, especially in cement
production.
The tube ball mills are steel rotary cylinders in which material breakdown is
carried out by the grinding bodies movement. Because of the cylinder’s rotation, the
unit formed by the grinding elements and the material is optimized for the grinding
action. Grinding is carried out by the collision and friction between the grinding
bodies and the shielded mill walls.
All the grinding systems must be operated in such a way as to obtain the
required production, as well as the necessary product quality, especially considering
the relevant production costs. Despite the numerous experiences in ball mill
operation, they are not often operated to maximum efficiency. The possible reasons
are not limited solely to inadequate composition of the ball load or to the wrong
position of the separator, but they also include a series of interrelated and globally
responsible factors of good system operation or its malfunction.
If the large number of possible combinations between variables is considered
in a grinding system, it is evidently hard to establish general instructions as to how to
operate and maintain it.
In the cement industry, close to 75% of the total electrical energy
consumption corresponds to the raw meal and clinker grinding system operations
and those for solid fuel grinding, when applicable.
Due to the diminishment of the available energy resources and the
uncontrolled rise of energy costs, it is important to pay more attention to reducing the
amount of energy used in grinding systems.
Material grinding inevitably generates substantial energy losses. The current
energy required for material reduction, given a certain particle size, greatly exceeds
the energy theoretically needed for its breakage until the corresponding increase in
At a very early stage in the history of mankind, certain instruments were used
for crushing food. Simple machines were developed to take the place of manual
labor. These machines were the ancestors of the blade mills driven by animals,
water or wind. Around the IX century, an increase in mineral deposit production led
to the wet process pounding mills, and in 1512 to the dry process impact mill was
developed.
the way for hard mineral primary reduction for raw materials. The cone crusher
made it appearance in 1877.
Table 6.11 summarizes the most important stage in this development.
Some time ago, there were drum mills with discontinuous or batchwise
operation available for fine grinding, in addition to the pounding mills and blade mills
and sometimes many of these were placed in mine shafts. Grinding cylinder were
made of ceramic materials or cast iron. Flint stones or ceramic balls were the main
grinding elements, and later cast iron balls were also used.
The dimensions of these first mills barely exceeded 1.0 m of diameter and 1.0
m in length. Therefore, they were all low-production equipment submitted to intense
work due to their cyclic method of operation.
But despite these disadvantages, the grinding process in drum mills was
really effective. Over the last quarter of the past century, great efforts have been
made to increase the industrial potential of continuous grinding.
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 21/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
The development of the tube mills (widely used at present for material fine
grinding) date back to over 100 years ago. The majority of the inventors started with
the assumption that proper sieves were needed for continuous grinding drum mills.
Therefore the coarse material obtained in sieving should then be returned to the mill.
In 1876, Gebruder Sachsenberg proposed a machine in which the sieving devices
were fastened to wooden frames surrounding the mill drum rotating with it. The
material that passed through the sieve made up the final product and was collected
on a deck which surrounded the machine. Whereas the material that was retained
in the sieve, was returned inside the mill through a special opening. This ball mill
discharged by sieves, still in use with practically the same design, can be operated
in wet or dry processes.
Its main disadvantage was that the size of the fine product depended on the
sieve mesh being used, which suffered frequent blocking and was subject to
extreme wear. These problems still occur today in operating ball mills with sieve
discharge.
For reasons of design, the whole drum was not formed as a single conical
drum, and it was divided into three conical chambers, as shown in figure 6.2 and
was proposed as the appropriate solution. However, these mills were not very
successful in practice because this arrangement of the mill tube also caused
considerable delays in material conveyance.
With the above in mind, the idea of speeding up mill feed conveyance by
means of a drop between the entry and exit ends of the mill was brought up. On
June 30, 1891, Konow and Davidson of Paris registered patent No. 62871 at the
imperial patent office of the German Reich under the title: “Ball mill with central
feeding and tangential discharge of the ground material.” This was to serve as the
basis of the modern ball mill.
The patent specification details are: “A ball mill is one which achieves a
continuous feeding operation process of the material to be ground through the
center of one end of the drum and discharging it in the perimeter of the other end of
the drum -this renders a constant difference in height between the entry and exit
points- producing a slow forward movement of the material being ground between
the balls in the longitudinal direction of the tubular drum which has sufficient length”.
v
v vv
Figure 6.2 First proposals for multi-chamber mills for continuous grinding.
Figure 6.3 shows an example of the design of this single chamber tube mill,
which could be supported on rollers or axles. The mill drum can also be conical in
order to increase the difference in height between inlet and outlet, as shown in figure
6.4. The object of the invention was related with the elimination of the deficiencies
associated with the operation of ball mills with discharge sieves normally used. This
means that the intention was to eliminate the main problem caused by the size of the
sieve that was operating and used to determine product fineness. Maintenance
costs were substantial and operation was hindered if low-density material was to be
finely ground. On the other hand, with the patented ball mill “the mill shell length was
selected so that the material to be ground would pass slowly along the entire length
of the mill at a progressive speed and be exposed for a long time to the grinding ball
effect”.
In this patent, the material to be ground was fed into the mill by means of a
screw or a vibrating conveyor. The material left the drum on the other end through
the holes or slots in the drum or the end wall. A discharge chamber was also
possible and there was temporary protection for the mill shell lining.
Equally interesting is the following focus of the patent under consideration: “If
the mill feed consists of small pieces, then one compartment is enough. However if
larger material is to be ground, several compartments with different size balls should
be used. In the latter case, the mill feed is first introduced into the compartment with
the largest balls and then it is passed from one compartment to the next; each
compartment with smaller sized balls than the previous one. These compartments
should be placed one under the other or along side each other.”
The Konow and Davidson patent expired on June 30, 1905 leaving the way
clear for other companies to build and develop improved tube mill models
unrestrictedly.
The first tube mill built in this century measured a maximum of 1,200 mm
diameter and 6,000 mm length. It was generally loaded with flint stones, and rarely
with steel balls. A simple lining was commonly used to protect the mill cylinder. The
productions ranged some 3 tph for cement kiln clinker grinding. Despite this
relatively low production, the tube mills for raw meal and clinker grinding in the
cement industry quickly became serious competition for the centrifugal tube mills
built in the United States by the Bradley company, and in Germany by Humboldt-
Maschinenbau AG, which had been until then the main type of mill used.
As time passed, there was a rapid increase in the tube mill dimensions used
both for wet-process grinding as well as dry-process grinding. Steel balls were more
widely used as a grinding means and the internal elements of the grinding chamber
were better adapted for a wider variety of operation conditions. Multi-chamber mill
were built and the closed circuit was introduced with the classification equipment.
Already in the XIX century, clinker from vertical kilns and rotary kilns was
ground in single-chamber tube mills. The size reduction was observed to take place
mainly in a length equivalent to 20-30% of the total mill length, whereas the
subsequent particle size reduction in the remainder of the mill was very small.
Consequently, the two-stage grinding system was developed in order to combine
pre-grinding and final grinding in two separate mills. The mill dimensions, that is, the
proportion of length and diameter, were selected according to the respective
grinding steps. Coarse material grinding was carried out in short, large diameter
mills; fine material grinding was done in long, reduced diameter mills.
In order to avoid using two transmissions for the separate mills, as needed in
two-stage grinding, a composite mill was introduced; coarse and fine material
grinding was combined in individual compartments in a single common mill. These
mill had from three to four individual compartment to satisfy the demands of product
fineness.
The composite mill, also called the open circuit mill, was and is successfully
applied for normal Portland cement production with a fineness of 2,800 a 3,500
cm2/g (Blaine).
There was a remarkable increase in mill dimensions between 1960 and 1970.
This had led today to the modern large tube mills with transmission mechanisms and
diameters over 6.0 meters for wet and dry process grinding. During this same
period, the autogenous (Aerofall) mill was installed in several plants for grinding wet
and binding materials.
Many other types of mills have been developed since 1980: roller vertical
mills, rollers presses, roller horizontal mills and others. However, the classic ball mill
is still one of the most important pieces of equipment in the cement industry, and
although its use is increasingly limited, it is still competitive at present.
Due a greater understanding of the grinding processes that take place inside
a ball mill, since the onset of the century the effectiveness of these processes was
recognized to depend critically on a la large number of parameters. Specifically:
Obtaining good grinding results with a low specific consumption of energy will
be possible only if these variables are selected correctly. In particular, the plating
configuration, mill rotation speed and the size and composition of the grinding
elements must be adapted to insure an optimal ball load movement en relation as
related to the feed material composition and the required product fineness. In
addition to the ball load motion, it is also necessary to make sure that the material
being ground has the necessary axial conveying speed (material standing time in the
mill).
Since over 70 years ago, it has been known that the mill linings are not only
indispensable to protect against wear, but also have a decisive technological
significance. The first ball mills were operated without special linings and efforts
were rapidly made to provide replaceable elements to protect the shell. At first,
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 27/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
these were made wedge shaped quartz or flint stone blocks, around 15 cm thick, cut
in a given form and drowned in cement. Later, they were long bars or borders in
which the mill feeding could be made. They were installed in the cylindrical shell
and used especially for mineral grinding mills. Figure 6.7 shows this type of
shielding for wet-process mills that were installed in some places even in the 20’s.
a b
c d
e f
Because ball mills are exposed to many different kinds and conditions of
operations, and the multiple specifications of the required product, one single type of
plating was found to be incapable of rending optimal results for every kind of
condition. Consequently, several types of shielding are still used, each with its own
advantages and disadvantages.
Using classifying shielding is one way to equal the forces on the feeding of a
single-chamber tube mill to the required fineness. These shielding produce an axial
motion of the grinding bodies due to the conical shape of the stepped plates. This
plating uses the centrifugal force, impact and slope effects to separate the large
balls toward the entry end of the mill and the smaller ball toward the outlet. The
starting point for the development of classifying shielding that began in the early
1920’s, was the Harding mill. In this mill, grinding body classification achieved by
the conical configuration of the drum. The relations is clearly illustrated in figure
6.8.
c
Figure 6.8 Development of classifying shielding based on the Harding conical mill.
a).- Carman classifying shielding.
b).- Slegten-Magotteaux classifying shielding.
c).- F. L. Smidth classifying shielding.
Some authors have tried to explain lower specific energy consumption in wet-
process grinding and have reached the conclusion that the particle has greater
mobility in an aqueous medium and can be ground more easily between the balls.
There is greater wear on the grinding bodies and the mill shielding or plating
in wet-process grinding and a high calorific energy consumption in the kiln. This is
the reason why this type of grinding is becoming less and less frequent.that the
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 30/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
recirculation pump supplies the oil at mill start-up. There must therefore be an
electrical interblocking system between the pump motor and that of the mill. When
the supports have high pressure pumps, they inject oil under pressure between the
smooth parts that are in contact immediately before start-up so that there will be a oil
film between them. The high pressure oil is supplied through a hole located in the
center of the bearing. The lubricant suspends the axle and thereby reduces friction
between the two metallic surfaces to a bearable degree. Once the mill is operating
normally, the high pressure pump is usually shut off. Lubricating oil should always be
kept a viscosity over 6° Engler.
Support temperature must be permanently monitored. Therefore, the sensor
should be in contact with the anti-friction metal. Measuring and controlling the oil
temperature is not considered appropriate because the lubricant’s poor thermal
conductivity leads to errors. For instance, it is possible that the metal has reached a
dangerously high temperature before the oil has gotten to it. For grinding with a
drying process, the supports are cooled by water. The minimum amount of water is
around 20-25 liters per minute for each support, at a maximum allowable pressure of
2 kg/cm2. However, this flow may not be enough if the temperature of the material
and/or gases passing through the support is too high, or if the water temperature is
high. The water flow should be started before starting up the mill, and should not be
shut off when the mill stops until the supports have cooled off completely.
Otherwise, the oil film between the contact parts will be squeezed before the axle
has completed its contraction, thereby causing the wear parts to scratch against
each other.
When bearings are manufactured, the contact surface is polished to a
roughness, “Ra'“ no greater than 0.8 microns. This roughness can be understood as
the arithmetical mean of the deviations of each point of the plate, “Yi”, with respect to
the average point of the same profile “Lm”. Figure 23 is a graphic representation of
the plate profile in the bearing, and roughness identification.
Surface
Height
In order to obtain a lubricating film between the axle and the bearing, these
contact surfaces must have different diameters. So, when the mill has been
shutdown, a wedge of lubricating oil is formed on each side of the point of contact.
During operation, this space is also filled with oil. It is also important that the bearing
does not grab the axle tightening it. If this happens because of factory defects in the
plates, the bearing edges should be filed down before mounting them. Figure 6.27
present the pressure profile. Angle “E” formed by the load direction, “P”, and the line
that passes through the axle and bearing centers, marks off the sector in which the
oil film is the thinnest. Just before the film’s minimum point, pressure reaches its
maximum value. The maximum load is located longitudinally in the center. The
supports are designed so that the bearing can withstand a pressure of up to 15-20
kg/cm2 if it is made of babbit metal, or 25-28 kg/cm2 if it is made of bronze.
Sliding shoes are water-cooled. The cooling water inlet is located at the
bottom point of the shoes.
The advantages of the shoe supports over the axle/bearing supports are:
x The bearings are eliminated, which in some cases have caused major damage.
x By eliminating the axles, the mill becomes shorter. Since the shoe construction
does not strangle the free section of the mill (as in the case of axle/bearing), there
is no obstruction for hot gas passage.
x Given the shorter distance between the supports, the mechanical requirements of
the shell are fewer and therefore the plating thickness can be reduced.
Mill
Ring Gear
Pinion
Drive
The lubrication for these drives is carried out by means of oil injection, with
lubricant cooling in an oil-water exchanger. Lubrication is affected by the peripheral
speed of the ring gear, the pressure between the gear teeth and roughness between
the teeth flanks. The relative movements between the pinion and ring gear axis
also have a negative effect on the lubricant conditions, because these movements
cause overloads increasing the pressure between the teeth.
Pinion
Mill
Ring Gear
Pinion
Drive
The lubrication for these drives is carried out by means of oil injection, with
lubricant cooling in an oil-water exchanger. Lubrication is affected by the peripheral
speed of the ring gear, the pressure between the gear teeth and roughness between
the teeth flanks. The relative movements between the pinion and ring gear axis
also have a negative effect on the lubricant conditions, because these movements
cause overloads increasing the pressure between the teeth.
Mill
Drive / Reducer
The low-speed axis of the reducer (twist axis) can be adapted directly to the
mill. The high-speed axis (coupled to the motor and/or clutch) has a pinion geared
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 36/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
to two intermediate wheels, placed on two intermediate axes, whose teeth gear to a
compensating wheel. The transmitted power is thus distributed among two
intermediate wheels and two intermediate axes, which enables the Symmetro
reducer’s compact construction. Additionally, due to the Symmetro reducer’s
symmetrical structure, the high speed axis and the twist axis are arranged so that
one is an extension of the other.
the low-speed axis of the reducer (twist axis) can be adapted directly to the
mill. The high-speed axis (coupled to the motor and/or clutch) has a pinion geared
to two intermediate wheels, placed on two intermediate axes, whose teeth gear to a
compensating wheel. The transmitted power is thus distributed among two
intermediate wheels and two intermediate axes, which enables the Symmetro
reducer’s compact construction. Additionally, due to the Symmetro reducer’s
symmetrical structure, the high speed axis and the twist axis are arranged so that
one is an extension of the other. The flexible suspension of the compensating wheel,
on membranes and disc springs, connected to the two guide rings, automatically
maintains equal force on the pressure of the teeth of the two gears of the
compensating wheel. This insures that each of the intermediate axes will transmit
half of the total power.
Another central drive is the satellite reducer type, (see figure 6.43). It is made
up of:
x Solar wheel and satellite pinions which are cemented, tempered and rectified,
which gear to the internal ring rear.
x Satellite holder held by bearing.
The three planet and solar pinion system, centered by means of the three
gears, without support, insure a secure and regular distribution of the spinning
momentum. The reducer temperature is automatically and optimally controlled by
an oil circulation system that serves the lubrication with respect to the outside
temperature.
The satellite reducer, made with hardened, polished and corrected dentate
both in the solar wheel and in the planets, has a performance of 98.9% at full load.
The axially coupled machine system, allows thermal dilation without hull
deformations or deterioration of the teeth flank contacts.
Motor
Drive/Reducer
Mill
Slotted plates
Lifting pallets
Inlet wall
(with slotted plates)
Outlet wall
(plating)
As mentioned previously, the vanes or pallets lift the material and ease its
conveyance to the next chamber which influences the material level in the previous
chamber. These pallets can be either radial or spiral. The longer the pallets, the
larger the amount of material can be lifted and conveyed. Consequently, the material
level in the previous chamber will be lower than in the case of shorter pallets leaving
ground material unlifted and unconveyed. When there are no lifting pallets, the
material level is at its highest in the previous chamber, and conveyance to the
following chamber is achieved by overflowing at the outlet wall. At mill start-up, the
pallet length is adjusted to the material level needed to obtain the best grinding
results for a certain product type and quality. Some modern diaphragms (Christian
Pfeiffer) regulate material passage by means of the outlet opening to the following
chamber. This prevents having to work on the pallet position which can pose a
problem as to its location and access.
When a diaphragm with long, axially full pallets (from anterior wall to posterior
wall) is used, the conveyance capacity is so high that the material level drops and
mill efficiency diminishes. In another place, a diaphragm without lifting pallets
(Sonex by F.L. Smidth) was introduced which led to a high material level because
the material has to pass over a high retention ring (outlet wall). These diaphragms
feature good grinding economy and low noise levels. However, due to the high
material level in the mill, energy consumption is low. Consequently, larger mills
have to be manufactured in order to obtain the same power consumption. A midpoint
between these two types is the short-pallet diaphragm (Stanex by F.L. Smidth),
which enables to keep an almost constant material level in the chambers and high
enough to insure proper mill efficiency without a reduction in energy consumption.
Shielding is the internal lining of the mill, made up of steel plates which
protect the shell from ball impact. Below are some of the other grinding functions
that shielding features:
x Lifts the balls as the mill turns for the impact or attrition grinding effect on the
material, depending on the degree or level of lifting; or
x Classifies the balls inside the mill, keeping a proper ball distribution (by size) for
material grinding along the mill.
These plates can be fastened to the shell in several ways.
Classifying
D ragpeb
Stepped
R ippled
Classifying shields are used to distribute the ball load in the mill, keeping the
large ball toward the inlet and the small ones toward the outlet. When used in a
single chamber, there is wide granulometry in the ball load. Along the mill, there is a
gradual lessening of the ball unitary weight, so that a two-chamber effect is attained
in a single chamber. Consequently, grinding efficiency is better than if the load were
a mixture of large and small balls. Additionally, the small balls will not be damaged
by impact with the larger ones, which would cause excessive wear. Classifying
shields are steps, axially oriented, mounted with the inclination facing the mill inlet.
When the ball strikes this shield, it tends to move toward the back (against the
material flow). Therefore, the larger the ball, the greater the impact and it will tend to
be thrown toward the inlet. See figure 6.37.
Material Flow
Classifying Shield
Mounted Shielding
Screw-on Shielding
Wedged Shielding
In the first case, each shield plate is separately screwed onto the shell. There
is a hole in the center of the plate for the fastening screw. Wedged shields consist
of rows of plates screwed onto the shell (as the previous ones) not consecutively.
The intermediate row(s) is(are) wedged between the ones that were screwed on.
Mount shields have lateral grooves or canals to fasten onto edges welded or
screwed onto the shell (similar to the installation of refractory brick in the kiln).
S WiBond
B k Fk80 3
Cs D
where:
k= Constant (36 for steel balls).
Fk80 = Size of material fed, expressed as the opening of a sieve that retains 20%,
en mm.
S= Actual density of material to be ground, en g/cm3.
WiBond = Bond grindability index; in other words, the specific energy needed to
grind the feed to a fineness of 80%, passing through a sieve of 100
microns. The index is expressed in kWh/t.
Cs = Mill speed, as a percentage of critical speed.
D= Effective diameter of the mill, m.
For the first chamber, the Fk80 of the freshly fed material granulometry (without
the returns) is considered. Normally, for the second chamber, the fact that the
material enter with a 20% residue of particles larger than 0.5 mm is taken into
account.
Another formula to determine the maximum ball size is the Papadakis
formula. Based on the result obtained in a test mill, the corresponding result is
scaled for a larger mill. It assumes that the ratio between kinetic energy of the ball
and maximum ground material size, is constant. Kinetic energy is determined from
weight (~B3) and the height of the fall (~D).
B 31 D 1 B 32 D 2
2 2
F1k100 F2k100
Using this as a basis, the maximum ball diameter for the larger mill is
calculated as follows:
2
F2k100 D
B2 3
2 1 B 31
F1k100 D 2
The minimum ball size is practically ruled by the size of the slots of the
intermediate diaphragm or outlet, depending on the load of the chamber in question.
However, Kassatkin calculate the minimum size with the following formula:
V2
Bmín Fk100
1.28 E Ubola D
where:
Bmín = Minimum ball size, cm.
Fk100 = Maximum feed size, cm.
V= Compression resistance of the ground material, kg/cm2.
E = Elasticity module of the ground material, kg/cm2.
Ubola = Actual ball density, g/cm3.
D = Effective mill diameter, cm.
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 42/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
In order to replace the balls keeping the load constant, it can be assumed that
ball wear is constant in a radial sense, regardless of its size. To illustrate the balance
load to constant wear, a simplified procedure is used, considering a 100-ball load
with consecutive sizes from 1 to 100 mm. After a certain operation time, during
which there has been a 0.5 mm wear to each ball, the result will be that the largest
ball now has a diameter of 99 mm (originally 100 mm); and that the smallest ball in
the original load (1 mm diameter) has disappeared. Thus, in order to have the
original ball load distribution again, it is enough to add another 100 mm ball.
The calculations to obtain the unitary weight (i) and specific surface (o), are
made for the largest fraction of the ball sample, taken as indicated in section 1.3 Ball
Sampling Techniques. It is in this fraction that the balls have a diameter that is larger
than half the nominal maximum diameter obtained by the Bond formula for maximum
ball size; that is, below this size, the ball’s effect on efficient grinding is null.
The weight of one ball (in grams) can be calculated in function to its
dimensions (diameter in cm) and its density (in g/cm3):
i 4.084 d3
However, the ball samples are weighed and counted in actual situations, The
resulting unitary weight is a quotient of the total ball weight divided by the total
number of balls.
The specific surface of one ball is calculated developing the following formula:
Surface S d2
o
Weight S 3
d U
6
6
o
d U
to obtain the specific surface with respect to unitary size (i), considering the density
of the iron at 7.8 g/cm3, the following is obtained:
1.23
o 3
i
The calculation for surface is based on the average unitary weight of the
grinding bodies and it supposes that all the grinding bodies are spherical.
where:
A = Number of balls, sizes between x and x+'x.
a = Replacement rate with diameter D balls, expressed in ball/sec.
s = Radial wear rate of the balls, expressed in cm/sec.
To find out the weight of these pieces, it is enough to multiply them times
each one’s unitary weight and density. The expression is:
Q ³ 6S x U 2a s dx
3
which, upon developing the expression of a fraction of the load, that is for balls
smaller than size d:
d
S U a 3
Q0
d
12 s 0
³
x dx
d S U a d4
Q0
12 s 4
Likewise, for the full load, that is for all the balls with sizes from 0 to D:
D S U a D4
Q0
12 s 4
The quotient of these 2 expression is the fraction of the balls smaller than a
certain size (d) with respect to the overall load:
d
Q0 d4
D
Q0 D4
Given the above, the weight of the balls that are smaller than half the
maximum size (0.5D) is only 6%.
When the balls that are smaller than a certain size (m) are separated from the
load, their distribution would be the result of integrating the previous expression from
m to D, and not from 0 to D.
D S U a D4 m 4
Qm
12 s 4
x Unit weight.
To calculate the average unit weight of the ball in a balanced load, the total
load weight is divided by the number of balls in the total load. Therefore, using the
previous expressions, the calculation will be as follows:
D
A0
D
³ 2a s dx
0
a
2s
D
D S U a D4
Q0 48 s
i D
A0 a D
2s
S U D3
i
24
in other words, one fourth of the weight of the balls used to compensate for wear.
If the balls smaller than size m are separated, unit weight is calculated as
follows:
D S U a (D 4 m4 )
Qm 48 s
i D
Am a (D - m)
2s
S U (D 4 m4 )
i
24 (D m)
x Specific surface.
Specific surface is also calculated from the same expressions mentioned above. In
these cases, we start with the calculation for the ball load surface (from 0 to D):
D
a S a D3
S0
D
³
0
S x2
2s
dx
2s 3
D S a D3
S0 6s
o D
Q0 S U a D4
48 s
8
o
UD
In the case of separating the balls smaller than m, the specific surface is
calculated with the following formulas:
D S a (D 3 m3 )
Sm 6s
o D 4 4
Qm S U a (D m )
48 s
8 D 3 m3
o
U D 4 m4
6.4.8 MATERIALS
The life span of mill parts (shields, plates and balls) depends on mechanical
and operation factors, but also to a great degree, on the construction material.
Technological developments in metallurgy have broadened the range of materials
used in making mill parts, improving their mechanical properties.
The following tables show some of the features of the materials used in
manufacturing balls for mills.
Raw meal,
soft,
Tempered 600 - 700 slightly Large ball 10 - 20 10 - 20
Wrought abrasive &
materials small ball 2-8 2 - 10
Raw meal,
Babbit 400 cement, Small ball 2-6 5 - 15
abrasive
steel materials
Raw meal, Large ball 2-5 2-5
Alloyed steel 600 - 700 cement, &
abrasive small ball
materials
1-3 1-3
The third, production capacity, with respect the first two aspects and the
available technology, determine energy consumption and the dimensioning aspects
of the mill (and peripheral equipment). Energy consumption is a function of the
amount of material ground, its grinding ease, feed granulometry and the desired
granulometry in the product:
The market has demanded finer products over the years; that is why energy
consumption has increased.
Similarly, there are standard mills for raw meal and coal. There are several
combinations of motor size, diameter and length. However, experience has shown
that the proportion of length/diameter adheres to the following quantities, in a closed
circuit:
Cement mills L/D = 2.6 to 3.8
Air-swept raw meal mills L/D = 1.5 to 2.0
2-chamber raw meal mills L/D = 2.0 to 2.5
3-chamber raw meal mills L/D = 1.7 to 2.2
Raw meal mills with central discharge L/D = 2.1 to 2.7
Open circuit mills have a L/D ratio of 3.0 to 6.0. When the L/D ratio
diminishes, the circulating load increases.
The intermediate diaphragm location is found by evaluating the specific
energy consumption and filling degrees, per chamber. Finally, the one closest to the
product parameters in chamber 1 is selected. These parameters are: 15% retention
in 0.5 mm, 7% retention in 1.0 mm, and 2% retention in 2.0 mm; and a 10 kWh/ton
consumption in the same chamber, approximately, for “normal” filling degree values.
Initially, the maximum ball size for chamber 1 is calculated, as seen in section
6.4.7.1 Maximum Ball Size, with the fresh feed; for chamber 2, all the material
coming from chamber 1 is considered.
On the other hand, the maximum ball size for chamber 2 depends on the
outlet diaphragm slot width.
The minim size should be less than the balance load obtained by maximum
ball size already calculated and/or, at least, twice the width of the aforementioned
slots.
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 53/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
Sometimes, especially when the mill feed granulometry covers a wide range,
2 or more different balance loads in diverse proportions can be combined: for
example, a load with a maximum ball size of 90 mm with another of 80 mm in
proportions of 50%-50%, 30%-70%, etc. This adjustment is usually made by a
specialist during the mill start-up test and during optimization.
Single chamber mills feature two (or more) “combined” balance loads which
are separated by the classifying shielding (they have no intermediate diaphragm). In
this case, each ball load is calculated from “imaginary longitudes” of each “chamber”
(imaginary location of an intermediate diaphragm) with the same filling degree. The
longitude ratios turn out to be from 40/60 to 45/55, approximately. Energy
consumptions for raw meal are: 5-6 kWh/ton in “chamber 1” (large ball) and 6-7
kWh/ton in “chamber 2” (small ball). The commonly used ball loads are:
x For coal grinding, 50 mm to 25 mm (or 15 mm).
x For raw meal grinding, 90 mm to 25 mm.
x For cement grinding (with pre-grinding), 50 mm to 15 mm.
Most solids have certain parts that are weaker than other due to such defects
as: dislocations, microcracks, borders between crystals and impurities. By applying
and external force, stress concentrates on the deformation points, thereby starting
cracks in the material.
By means of a balance of forces in the mobile end of the crack, it has been
proven that stress must reach a critical value to initiate breakage. Once breakage
has started, of the deformation force is not quickly relieved, the system becomes
unstable and cracking is accelerated to up to 40% of the speed of sound within the
solid.
In this process, energy liberated exceeds the required value, so that
remaining energy can concentrate on other cracks and spread. Once the crack
starts to spread, fractures can result.
In materials with a high rigidity index, the energy applied by an external force
is released as the propagation of deforming cracks, making the material easier to
break.
On the other hand, the probability of breakage of a particle increases with the
number of forces applied, up to a number in the order of ten. If more of these are
applied, the probability of breakage decreases and the probability increases that the
cracks will only run near the periphery of the solid.
This is of special interest in the case of mills that work with a bed of material,
where it is very important to eliminate the fine particles because otherwise the
probability of larger particle breakage is reduced by the increase in contacts due to
the small particles.
The force application speed is more important for viscoelastic materials than
for rigid ones. Viscoelastic materials are fractured more easily at high speeds. This
is why plastic materials are better ground by impact than by slow compression.
As to particle size, small particles have a different internal stress pattern from
the large particles, mainly because of imperfections, stored elastic energy and
plastic deformation. In general, this is what makes them harder to break. The
smaller the particles, the fewer the imperfections. Therefore, stress levels have to
be increased accordingly.
The second reason why small particles are harder to break is that the elastic
energy stored at the moment of starting the crack decreases proportionally to
volume, while the energy needed to spread the crack decreases proportionally to the
transversal area. Therefore, when reducing the size of the particle, it may be that
energy stored and released at the moment of initiating the crack might be insufficient
for propagation, and so the crack stops spreading.
The effect of plastic deformation becomes more important as the particle size
decreases. For example, with average limestone, size reduction of up to 5 microns
increases the number of radial cracks. Under 5 microns, plastic deformation
increases so that there is no increase in radial cracks. Finally, at 1 micron the
particle only deforms without generating cracks, in this case, shows a totally plastic
behavior. Studies with cement clinker have established that the transition from a
rigid behavior to a plastic one appears at between 10 and 20 microns.
Lastly, it must be pointed out that grinding additives have no effect on the
stress patterns in the particle, nor do they help to the propagation of cracks or
particle fracture. Their positive effect for industrial grinding resides mainly in the fact
that they confer a greater motility to the particles, increasing the probability they will
be reached by the impact of one of the balls .
Clearly, actual grinding includes the application of forces to the greatest
possible amount of particles. Since many of the forces applied to particles do not
reach the required value for the development of cracks and fractures, much of the
energy will be lost as heat. Therefore, it is to be expected that industrial grinding
systems are highly inefficient.
100.00%
64 64
32
32 8
8 4
50.00%
1
4
1 Feed Size
32-42 mesh
60-80 mesh
0.00%
0 100 200 300 400 500
Different experiments have proven that the grinding system used has no
bearing on the distribution that results from initial breakage (bimodal distributions
occur in every case). As grinding time elapses, the distribution starts to turn into
monomodal, and the modal value of the smallest particle prevails.
For impact grinding, an increase in impact speed has the same effect on
particles distribution as increased grinding time in a ball mill. Therefore, all the
grinding methods render particle distributions which go through the same cycle of
changes as grinding progresses.
indicate that it is the energy accumulated in the particle at the moment of fracture,
more than the magnitude of the external energy applied, what really decides the
resulting particle distribution. Therefore, external energy in needed to start breakage
but not all of it is used in the fracture itself.
On the other hand, materials undergo a transition from a rigid to plastic state
when reduced below certain particle sizes. In cement this transition occurs at some
point between 10 and 20 microns. Any point at which this transition happens, could
be expected to be related with the grinding limit, since any further grinding action
would only change its shape, but not its size.
S ( m in - 1 )
11.0
0 .5
0 .2
0 .1
0 .0 5
0
0 50 100 200 500 1000 1
m ic r a s
Cataract
Cascade
Minimal
grinding zone
1. The balls rise with the shielding tend to “lock” together, depending on the degree
of sliding produced by the shielding. Therefore their grinding work is greatly
diminished. In this zone, the minimal grinding work that takes place is carried out
basically by attrition and cutting.
2. The balls that emerge from the inclined surface of the ball bed roll on the surface,
provoking many impacts in the material load. This movement is known as
cascade and is very effective for the fine grinding of small particles, depending
on ball size. In this movement the grinding work is carried out mainly by attrition,
cutting and compression.
3. At greater mill speeds, with a given load of balls and materials, part of the balls
are thrown up and away from the bed of balls and fall to the bottom of the mill.
This movement is called cataract and is very effective for the grinding of larger
lumps of material, such as in autogenous mills. In this zone, grinding work is
carried out mainly by impact.
The fraction of critical speed at which movement passes from cascade to
cataract varies with the degree of filling with balls and material. In actual conditions,
a speed must be found that approaches the desired optimal effect, both in chamber
1 (where cataract is mainly desired) as well as in chamber 2 (where cascade is
mainly sought). This speed will surely be close to the value where transition from
one to another is carried out (around 75% of critical speed for typical filling degrees).
x Ball Size.
Laboratory tests have found that for particle sizes below 1 mm, there is a reduction
in the function of breakage frequency when ball size is increased from 3/4” to 2”.
However, an increase in ball size also displaces the value of particle size at which
the greatest “S” occurs to larger values. These relationships can be seen in figure
6.45.
x Ball Density.
A proportional relationship has been found between the function of breakage
frequency and ball density; so the greater the ball density, the greater the breakage
frequency.
S
-1
(min)
1.0
2.0 in
1.5 in
1.25 in
1.0 in
0.75 in Ball Size
0.1
Particle Size
1.0 10 mm
Figure 6.45 Typical variation of “S” with the ball and particle sizes.
fc = 0.4 q U
Overfilling
0.6
begins
0.5
Constant S range
0.4
0.3
0.2
0.1
Interstitial filling U
Figure 6.46
However, ball size does have a striking effect, sensibly increasing the
generation of fine particles when ball size is increased.
c(t)
Function I(t) has the same shape as curve C, but in terms of percentages.
Also, from I(t) the average retention time can be calculated as:
f
W = ³ tI(t) dt
0
Of all the parameters and mill geometry, when considering industrial mills,
there are basically two that have importance in the RTD function: mill speed and ball
size.
Considerable changes have been found in the shape of the RTD function and
in the average time of retention when changes in the velocity of the mill cause a
modification in the movements of the balls from cascade to cataract. However, a
greater increase in speed within the cataract range does not have a considerable
effect.
On the other hand, when reducing ball size and reducing interstitial space
between the balls, the material passage area is reduced, increasing retention time
as long as U<1.0. For values of U>1.0 the difference in retention times begins to
decrease.
%
Gases from
Preheater %
Kw
%
Tgcal Pent Tsal
Tent Noise
Figure 6.48 Raw meal grinding setup with first generation separator.
Table 6.7 Main Control Parameters of a Typical Raw Meal Mill and their Influence on
the Rest of the Parameters.
Suction at Mill
Suction at Mill
Outgoing gas
temperature
temperature
Noise Level
Main Motor
VTI Motor
Indication
Outgoing
Entrance
moisture
fineness
Elevator
material
Returns
Product
Product
Power
Power
Power
Outlet
By raising:
Feed Ï Ð Ï Ð Ð Ï Ï Ð Ð Ï Ð
Moisture Ð Ð Ï Ð Ð Ï Ð Ð Ð Ï Ð
Inlet Temp. Ï Ð Ð Ï Ï Ð Ï Ï ÏÐ Ð Ï
VTI Blind Ï Ï Ð Ï Ï Ï Ï Ï Ï Ð Ð
Kw % %
%
tss
Kw
% Kw
Kw, rpm
Kiln
Gases
Feed: tph % H2O
% °C Fineness
Cold Air % H2O
% tsal
Returns
Psal
tph
Kw
Figure 6.49 Raw meal mill arrangement with a third generation separator.
Table 6 8 Main Control Parameters of a Raw Meal Mill, and their Corresponding
Influence on the Rest of the Parameters.
Outgoing material
Product moisture
Separator Power
VTI Motor Power
Product fineness
Separator Outlet
Temperature at
Elevator Power
Separator Fan
Suction at Mill
Suction at Mill
Outgoing gas
temperature
temperature
Noise Level
Main Motor
Indication
Entrance
Returns
Power
Power
Outlet
By raising:
Feed Ï Ð Ï Ð Ð Ï Ï Ð Ð Ï Ð Ð Ï Ï
Moisture Ð Ð Ï Ð Ð Ï Ð Ð Ð Ð Ð Ï Ð Ï
Inlet Temp. Ï Ï Ð Ï Ï Ð Ï Ï Ï Ï Ï Ï Ï Ð
Separator Speed Ï Ð Ï Ï Ï Ï Ï Ð Ð Ï Ð Ð Ð Ð
Separator Blind Ð Ð Ð Ï Ï Ð Ð Ï Ï Ð Ï Ï Ð Ð
Recirculation Blind Ð Ð Ï Ð Ð Ï Ð Ð Ð Ð Ð Ð Ð Ï
VTI Blind Ï Ï Ð Ï Ï Ï Ï Ï Ï Ï Ï Ð Ï Ð
Although the amount of returns, gas flow and temperature, feed moisture,
hardness and granulometry, and the conditions of the mill’s internal elements may
influence noise level, none of these conditions can be controlled with such a rapid
effect as the fresh feed, which is why it is precisely this last parameter that is the
primary control of the mill’s level.
If there is false air, it will have a great impact on the gas temperature at the
mill inlet.
%
%
Fresh Feed
°
tp To silo
% tp Ps
tsal
Table 6.9 Control Parameters for a Cement Mill equipped with a First Generation
Separator.
Outgoing material
Separator Power
VTI Motor Power
Product fineness
Elevator Power
Product Blaine
Suction at Mill
Suction at Mill
Outgoing gas
temperature
temperature
Noise Level
Main Motor
Indication
Entrance
Returns
Power
Outlet
By raising:
Feed Ï Ï Ï Ð Ð Ï Ï Ð Ð Ï Ð Ð
Moisture Ï Ï Ï Ð Ð Ï Ð Ï Ð Ï Ð Ð
Inlet Temp. Ð Ð Ð Ï Ï Ð Ð Ð Ï Ï Ï Ï
Static Separator Ð Ð Ð Ï Ï Ð Ð Ð Ï Ð Ð Ð
Dynamic Separator Ï Ï Ï Ð Ð Ï Ï Ï Ð Ï Ï Ï
Water Injection Ï Ï Ð Ð Ð Ð Ï Ï Ï Ï Ð Ð
% tss %
Kw
%
P collector
P collector
Kw
Kw
Kw, rpm
Fineness
Blaine
Returns
Feed: tph tph tsal
°C Psal
% H2O
Kw
%
Pent
lts/hr Noise tsalmat lts/hr
% tdiaf %
Table 6.10 Main control parameters for a Cement Mill equipped with a Third
Generation Separator, and how they influence the rest of the parameters.
Outgoing material
Separator Power
VTI Motor Power
Product fineness
Elevator Power
Product Blaine
Suction at Mill
Suction at Mill
Outgoing gas
temperature
temperature
Noise Level
Main Motor
Indication
Entrance
Returns
Power
Outlet
By raising:
Feed Ï Ð Ï Ð Ð Ï Ï Ï Ð Ï Ð Ð
Moisture Ï Ð Ï Ð Ð Ï Ï Ð Ð Ï Ð Ð
Inlet Temp. Ð Ï Ð Ï Ï Ï Ï Ï Ï Ï Ï Ï
Water Injection Ï Ï Ð Ð Ð Ï Ï Ï Ï Ï Ð Ð
Air flow to separator Ð Ï Ð ÏÐ ÏÐ Ð Ð Ï Ï Ð Ð Ð
Separator rotor speed Ï Ï Ï ÏÐ ÏÐ Ï Ï Ï Ï Ï Ï Ï
Mill draft Ð Ï Ð Ð Ð Ï Ï ÏÐ Ï Ð Ð Ð
The relative amount of material in the circuit can be obtained by the separator
amperage, the returns and the elevator power.
Coal mills are generally 100% air-swept and equipped with a static separator.
Kw Kw
%
%
tsal
Psal
O2
tph
% H20
% Satellite
% % Kiln
% Kw
Noise
tent
Pent
O2
If there is false air, it will have a great impact on the gas temperature at the
mill inlet.
good indication of the amount of material inside the mill, and the degree of stoppage
in the partition walls or diaphragms.
The source of heat can be kiln gases or heat generator gases. The only
difference between the two is the gas temperature.
Drying can be carried out in a drying chamber directly connected to the mill, in
a separator, or separately with no connection to the mill.
However, in all cases the objective is the same: to reduce the moisture
percentage in the feed to a level that the grinding chamber can grind the material
with no problems of sticking to the balls or the lining.
In other words, drying control implies keeping the adequate moisture level -
usually between 1% and 1.5%- in the material that enters the grinding chamber.
Sampling between the drying and grinding process is only possible when the
drying process is carried out in separators or flash. That’s why a ratio between the
remaining moisture and temperature in the material at the outlet (of the separator,
flash or mill) is usually made.
The higher the temperature, the faster the materials will dry, and the lower the
volume of gases needed, rendering lower air speeds and fewer pressure drops.
The limit of the amount of gas is actually the maximum air speed. That is, if
large amounts of gas are required, the equipment would increase in physical size so
as not to surpass the maximum air speed limit. In actual conditions, the air speed in
a mill or a separator will be higher when it’s necessary to change the mill air more
often to eliminate the moisture that has evaporated, or when conveyance inside the
mill has to be increased.
x Grinding.
Grinding is carried out in the grinding chamber(s) (usually 1 or 2). Then the
separator separates the mill product into fine and coarse particles. To control this
process and make it as cost-efficient as possible, the relations between the different
factors in the process have to be taken into consideration.
Theories indicate that, once the machine sizes are given, the composition and
amount of the ball loads, a certain amount of material with respect to the ball level
will off the most efficient grinding. These theories also indicate that a certain final
product fineness corresponds to an optimal circulation factor. In other words, in
order to obtain maximum production at the desired fineness, the proportion between
feed and returns to the separator has to be kept.
Most of these data are registered continually in the control panel. Only
fineness is the result of laboratory analysis every 1 or 2 hours.
Amount of feed.
Separator adjustments (pallets, blades, blinds, rotor speed or air flow).
Mill air flow adjustment.
Normally, the separator is adjusted first to get the desired product, then the
feed is adjusted to obtain the optimal recirculation factor. Later, it’s necessary to
maintain this balance, along with the normal variations of feed, such as:
granulometry, composition, hardness and density.
To do this, the noise level and the elevator energy consumption are the most
useful values. By keeping these two parameters constant, minor adjustments to the
feed are enough to obtain optimal conditions in the mill. Separator adjustments are
made only on the basis of the fineness results obtained.
Table 6.11 Content of volatile substances, moisture and fineness for different types
of coal.
Type of (%) (%) (%) Hygroscopic Required
Coal Volatile Subst. Ashes Water Fineness
Anthracite 5 5 2 5-7% + 90 P
Anthraciferous soft coal 5-15 3-5 2-6
Semi-oily coal 15-20 5-8 2-6
Bituminous coal 20-30 8-10 2-6
Oily coal 30-40 10-20 2-6
Lignite 40-50 15-30 10-25 30% + 90 P
cases the objective is the same: to reduce the moisture percentage in the feed to a
level that the grinding chamber can grind the material with no problems of sticking to
the balls or the lining.
The best way to control drying is by keeping a constant amount of material
entering the chamber. Outlet humidity will depend on the type of coal to grind. A
relation between the remaining humidity and material temperature at the mill outlet is
generally used.
Work should generally be carried out with the highest possible temperature,
considering the maximum temperatures the ducts, fans, axles or shoes and the
other equipment in contact with the gases can withstand.
The normal maximums are:
With kiln gases = 400°C
With generator gases = 800°C
The higher the temperature, the faster the materials will dry. Additionally, a
lower volume of gases will be needed, thereby rendering lower gas velocities and
pressure drops.
The actual limit to the amount of gases is the maximum air speed. In other
words, if large amounts of gases are required, the equipment will increase in size so
as not to go beyond the maximum air speed limits. In everyday practice, the mill or
separator gas velocity will be higher when it is necessary to change mill gases more
often to eliminate the evaporated humidity, or when conveyance within the mill
should be increased.
x Grinding.
Coal grinding is normally carried out in an “inert” environment to prevent
ignition or the risk of explosions. The word “inert” refers to a low oxygen percentage
environment in which there is no possibility of fire or explosion; usually one with less
than 10% of O2.
Operating an inert system is somewhat more complicated since it is
necessary to control the mill temperature, oxygen percentage, and the dedusting
filter temperature.
As it is not possible to lower gas temperatures with cold air (increasing the O2
percentage), a water-injection system is needed.
On the whole, the hot gas flow is kept constant and water injection adjusts the
output temperature depending on feed moisture.
In some cases, recirculation is also possible so as to restrict water injection.
Grinding is carried out in the grinding chamber(s) (usually 1 or 2).
Then the separator separates the mill product into fine and coarse particles. To
Chapter: 6 Written by: JEA, EMT 13/11/2002
Version: 1.0 Revised by: EDP 86/102
BALL MILLS
DIRECCIÓN TÉCNICA TRAINING MANUAL
control this process and make it as cost-efficient as possible, the relations between
the different factors in the process have to be taken into consideration.
Theories indicate that, once the machine sizes are given, the composition and
amount of the ball loads, a certain amount of material with respect to the ball level
will off the most efficient grinding. These theories also indicate that a certain final
product fineness corresponds to an optimal circulation factor. In other words, in
order to obtain maximum production at the desired fineness, the proportion between
feed and returns to the separator has to be kept.
Most of these data are registered continually in the control panel. Only
fineness is the result of laboratory analysis every 1 or 2 hours.
Theories indicate that, once the machine sizes are given, the composition and
amount of the ball loads, a certain amount of material with respect to the ball level
will off the most efficient grinding. These theories also indicate that a certain final
product fineness corresponds to an optimal circulation factor. In other words, in
order to obtain maximum production at the desired fineness, the proportion between
feed and returns to the separator has to be kept.
Most of these data are registered continually in the control panel. Only
fineness is the result of laboratory analysis every 1 or 2 hours.
Process Value ( PV )
Outlet (O)
PROCESS
Set Point ( SP ) Set Point ( SP ) MEASURING
Operator INSTRUMENT
C o n tr o la d o r P I
S U P ER V IS IÓ N
A M PL IFIC A D O R
O N / O FF
A L IM E N T A C IÓ N
TO TAL
M IC R Ó F O N O
S U P ER V IS IÓ N FOLÁFO NO
A M P L IF IC A D O R
A M PL IF IC A D O R A M PL IFIC A D O R
B
A A -B
ALARMAS
M M
A I
X N
I I
M M
O O
YE S O C L IN K E R
0 -10 % 0 -100 %
In other words, automatic control is an aid and a tool for the operator, but not
a substitute.
surface temperature of the entire system above the dew point to prevent water
condensation. This is done with a small amount of draft, with an outlet temperature
between 100- 120°C and it is used only in raw meal or coal grinding. The fan blinds
should be slightly open (between 5 and 10%).
5. The circulating load group is started up. The equipment operating in this group
includes the separator, the separator feeding airslide fan(s), the elevator(s) and the
mill discharge airslide.
6. Coupling or mill start-up. At this time, the mill is set into operation. In some
systems, the feed belt also begins to operate.
7. The feeding system begins operation. The feeding belt and scales operate in this
set. The mill should not operate empty for over 10 minutes. This is very important
so as to prevent damage to the shielding plates.
8. The fan blinds should be opened to provide mill sweeping and air to the
separator. The fuel regulating valve should be opened if there is a gas generator.
75-80% of the maximum feed should be fed manually, for a period of 10 to 15
minutes.
9. The next step is to stabilize the mill operation. It is during this stage that a great
deal of attention should be paid to the mill operation parameters and variables,
observing tendencies such as: inlet and outlet pressure, temperatures of incoming
and outgoing gases and solids, mill noise level; amperage and/or kilowatts of the mill
discharge elevators, the main motor, the sweeping fan and the separator.
10. Based on the tendencies of the aforementioned parameters, the next step is to
recover the mill feed gradually, in approximately 5% increments of the total feed at
10 to 15 minute interval, until 95% is reached. Then, feed is switched to automatic
so that it can reach the maximum feed. The necessary adjustments should also be
made to pressures and temperatures, and feed is increased up to its maximum
production.
For cement mills, it is necessary to wait until the system has warmed up, in
other words, that the intermediate diaphragm temperature should be above 105ºC in
order to proceed to start up the water injections pumps (if applicable), trying to keep
the temperature within the preset values (normally 110ºC at the intermediate
diaphragm and fro 115ºC to 120ºC at the outlet diaphragm).
The first product sample should only be taken as a guide and not as reference
to make any corrective actions to fineness or Blaine, unless the data is too far off
range.
Unexpected Shutdown:
An unexpected shutdown is one in which the operator does not intervene or if
he does, he has only enough time to uncouple; in other words, the time to stop the
mill is unknown. The shutdown may be due to process flow obstruction, equipment
failure that uncouples the mill, electrical energy supply failure, etc.
In a situation such as this, the following points should be considered:
1. Make sure that the heat source gas gate is closed, if there is drying. A cold air
gate should be opened if the temperature skyrocketed. Contrarily, in the case of a
coal mill, no gates or blinds should be opened so as to prevent oxygen from entering
the system. Cooling should be carried out by water injection.
2. If the system does not have interblocking to stop the water and additive injection
pumps, these should be manually stopped immediately.
6.8.3. INTERLOCKING
Interlocking refers to the requirements that must be met in order to operate
the equipment. any deviation in this aspect will hinder startup or cause the facility to
shutdown.
As mentioned previously, the number and types of interlockings will depend
on the specific facility in question. However, they all have the same objective: to
protect personnel and the facilities.
The process interlocking for a specific cement mill are given below, and the
startup sequence so as to insure material flow.
x The dosing scales have a signal (light and sound) on the control panel, in case of
belt misalignment or lack of material.
x The additive pump has a light and sound signal, in case of additive flow failures.
x The lubrication pumps for the main bearings, main reducers and the mill crown
have a light and sound signal on the control panel which goes off in case of oil
flow failure. If the failure is not corrected in 5 minutes, the mill will shutdown.
x The air compressor of the crown lubrication pump emits a light and sound signal
in the control panel when there is air pressure failure. The mill will shutdown if it
is not restored within 5 minutes.
x The mill drive temperature measuring points set off a light and sound warning
signal when the temperature is 75°C. If the temperature is not controlled and
reaches 80°C, the mill will shutdown.
x The separator reducer lubrication pumps have a light and sound signal which
stops the separator when there is a flow failure.
x Both the mill bag filters as well as that of the separator set off a light and sound
signal when there is air pressure failure. The mill shuts down if the mill bag filter
fan stops.
x The aeropol has a light signal on the control board. In case of overfilling, the feed
to the mill is automatically stopped until the aeropol load goes back to normal.
x Electrical motors are protected against overloading. In case of an overload, the
equipment is triggered setting off the light and sound signal on the control panel.
x The mill outlet temperature is interlocked with the cooling water pump. If the
temperature is below 105°C, the pump stops. When it reaches 115°C, the water
pump begins to operate.
The startup sequence for a specific cement mill is given below:
1 Clinker and/or gypsum scales.
a) Scale master control.
b) Additive pump.
c) Cooling water pump.
d) Mill.
2. Mill.
a) Main reducer’s lubrication pump.
b) Main bearings lubrication pumps
c) Main bearing lifting pumps.
d) Mill gear wheel grease pump.
e) Air pressure failure.
f) Failure due to high temperatures at measuring points (T maximum = 80ºC).
g) Airslide.
h) Bucket elevator.
3. Bucket elevator:
a) Airslide.
4. Turboseparators.
5. Turboseparator dispersion plate:
a) Turboseparator motor fan.
b) Turboseparator lubrication pump.
c) Bag filter extractor.
6. Bag filter extractor:
a) Shaking program.
b) Shaking turbo fan.
c) Hopper worm gears.
d) Cyclone airlocks.
e) Cyclone worm gears.
f) Returns airslide.
g) Airslide fluidizer.
7. Fluidizer from airslide to aeropol.
8. Aeropol compressors.
9. Fluidizer from airslide to silo.
If the interlocking on a piece of equipment numbered from 1 to 9 breaks open,
all the rest of the equipment in descending order will also snap, except the
equipment with an independent connection and the mill lubrications pumps.
Quality
Bottle necks. (Evaluate peripheral equipment)
Mechanical evaluation. (Related to process)
Tools:
Updated historical data
Internal inspection of the mill
Ball load sampling
Axial sampling
Matter and heat balance
Separator analysis
% Ret
80
Tamaño de Partícula
Figure 6.58 Particle size vs. Retention Percentage.
To calculate the maximum ball size, we’ll use K80 considering the overall
distribution. It is preferable to calculate the maximum size per component, or it
that’s not possible, the maximum size for overall feed.
There are two possible approached in calculating maximum ball size:
1. The ball size needed to grind the material.
Additionally, if the kiln has a grate cooler, it is advisable not to cool the clinker
too much, and convey it hotter to the mill.
When there are problems to dry the material in the mill, the draft should be
lowered taking care always not to approach the dew point.
Another factor that can help it to change de circulation factor, because if the
returns return to the mill cooler than when they left, they are drawing heat from the
system, and this reduced drying.
When there is excess moisture in one of the materials, contact for long
periods of time before entering the mill should be avoided because this may cause
premature clinker hydration.
x Temperature:
The easiest method for controlling temperature in the mill is by water
spraying; it is more efficient than air cooling. The quality of the cooling water is
important because if it has solid particles, it will cause excessive wear to the
spraying nozzle. It is also important to make sure the water has no organic matter
(such as river water) that could affect product resistances.
Care should be taken when putting gypsum in contact with hot clinker
for long transport distances, because it could cause dehydration. Some gypsums
are more sensitive than others.
x Chemical Composition:
Chemical composition has more to do with product quality than will the
grinding process.
% R +45 microns
12 12
10 10
% R +90 microns
8 Í KILN 8
6 6
4 CALCINING Î 4
2 ZONE 2
0 0
0 2 4 6 8 10 12 14 16 18 20
Volatile content %
The ash content in the coal at the moment of designing the raw mixture is
also something to watch carefully as the ash contain silica which could affect clinker
quality.
Some raw materials contain coal, so special care should be taken because
they start to burn in the first stages of the pre-calcining zones and may cause
scabbing.
IN cement grinding, the different clinker components vary greatly as to
grindability. The C2S is harder to grind than the C3S, so grinding is more difficult
when C2S increases.
If clinker is weathered, it’s harder to grind.
chambers in chamber 1 –one very short load with a coarse balance load and the
other longer with a balance load of maximum 60-70 mm.
To correctly assess ball performance by axial material sampling, the
circulation factor should always be considered. When the circulation factors are
high, the grinding effect is diminished because of the large amount of ground
material that goes back with the returns.
When axial sampling shows an increase in the retained material in a single
point, the circulation factor should not be considers, unless this happens in tow
consecutive samplings. If the phenomenon is accumulative, a single larger sized
particle that is detected will impact the result in the rest of the sieves.
6 5 4 3 2 1 1 1
1.0 %
10.0 %
50.0 %
90.0 %
95.0 %
97.0 %
99.0 %
1 mic 10 mic 100 mic 1 mm 10 mm 100 mm
It is sometimes hard to justify the use of a third generation separator just for
energy saving, as the following benefits should also be considered:
x Ease of changing the product.
x Increased cooling/drying capacity.
x Product quality stability.
x Elimination of scrap. (Not possible in all third generation separators.)
REFERENCES
x Blue Circle Industries PLC. Mill Design Principles, Milling Systems. Cement
Technology, Volume 2, 1979
x C. L. Prasher. The Physical of Breakage, Particle Shape, Size and Surface.
Crushing and Grinding Process Handbook., 1987
x F. L. Smidth. Grinding Mills. Compendium for Cement Production Seminar, 1987.
x F. L. Smidth. Grinding, Raw Meal Mill process Control. Production, Quality Control
and Process Course. 1987.
x F. L. Smidth. Symetro Reducer, type TSX, Maneuvers and Maintenance.
Instructions 33844-02.
x F. L. Smidth. Support with Sliding Shoes and Inclined-Base Plates. Instructions
27084-14.
x Holderbank. Cement Grinding. Cement Seminar, Process Technology 1, Quito,
Ecuador. 1986.
x Labahn/kohlhaas. Cement Grinding. Cement Handbook. 1985
x W.Schiebe, Freiberg. 100 Years of Tube Ball Mills. A Historical Review. Germany,
ZKG 3/1993, pp. 123-128
x Walter H. Duda. Grinding in Cement Production. Technological Manual on
Cement. 1977