VALID PRACTICE

FOR THE
PRODUCTION PROCESS

#10 – BALL CEMENT MILL

MONITORING, INSPECTION & EVALUATION

Athens, December 2007

 Valid Practice #10

16/10/2007 Page 2 of 20
 Valid Practice #10

BALL CEMENT MILL

MONITORING, INSPECTION & EVALUATION
1.1. Grinding efficiency
Step
Inlet Linin Intermediate Classifying
First Diaphragm Second Outlet
Trunnion g Lining
Chamber Chamber Trunnion
Plates Plates

Fine Outlet
Mill End Coarse Diaphragm
Grinding
Wall Grinding Media
Media

Maintaining the condition and effectiveness of the equipment relies on regular site
inspections and measurements around the system. Internal inspection of the mill
should take place on regular basis to assess the condition of the charge and the
liners. Periodically axial samples need to be taken in the mill to assess the
progression of fineness development along the grinding path. From this the
appropriateness of the media size grading for the fineness of the predominant
cement produced on the mill can be assessed. Samples also need to be taken
around the milling circuit to assess the recirculation load and the characteristics
and efficiency of the separator. Temperatures and pressures around the milling
system need to be measured to calculate air velocities and a mass and energy
balance. In combination this information will provide a comprehensive
assessment of the condition of the milling equipment. Periodic gathering of this
information is useful for detecting changes in the condition and the performance
of the mill.
Each ball mill should be inspected at minimum every month. Along with the
internal mill inspections, the ball charge should be monitored according to the mill
kilowatts. The mill should be inspected as soon as mill kW drops by 5% from the
targeted mill kW. During each inspection, a “Monthly mill ball charge inspection
sheet” should be filled out.

16/10/2007 Page 3 of 20
 Valid Practice #10

1.1.1. Guideline for Crash Stop

The CM should be crash stopped and visually inspected internally.
i. All associated departments should be informed at least 24hours prior to crash
stopping the mill. The departments involved include Production, Maintenance,
Process and the Quality Control.
ii. The CM must be crash stopped while the mill is at maximum production and
has been running stable for at least 4hours to give consistent representative
samples.
iii. The control room operators will crash stop the mill by stopping the mill and all
related equipment including all fans and bucket elevators. If possible, the mill
is stopped so that it can be entered without any further turning. It must be
ensured that material feeding is not stopped previously and that also the mill
fan stops simultaneously to prevent that fines are discharged with the air flow.
The product conveying equipment should remain running to convey the
remaining product from the lines and hoppers.
iv. The mill will be prepared for inspection according to the respective safety
procedure. Estimated time for preparations is 2hours. It should be noted that
at all time the fuses should be removed from the respective motors and that
nobody should enter without clearance from the Production manager and the
Maintenance Manager.
v. The mill is allowed to cool down. Estimated time for cooling is 5hours.
vi. To check the grinding progress (grinding efficiency over the compartment
length), axial sampling is performed.

The procedure is as following:

 Take samples from below surface, below level of balls, along the mill axis.
 Assess presence of “nibs” (e.g. coarse material, ~5mm) especially near the
outlet of the 2nd compartment.
 The most important (“Key”) samples are those on either side of intermediate
diaphragm as well as at mill outlet.
 Usually 3 samples in chamber 1 and 5 samples in chamber 2 are collected,
whereas in theory we should collect axial samples after every 1m from the
mill inlet to the intermediate diaphragm as well as from the intermediate
diaphragm to the mill outlet and at each point we take a set of three samples
along the mill chord.

16/10/2007 Page 4 of 20
 Valid Practice #10

Sample

The result shows whether or not the grinding progress in the mill is good.
Furthermore, axial sampling can provide indications of possible causes in case of
disturbances.
A typical example of axial sampling is presented hereafter:

16/10/2007 Page 5 of 20
 Valid Practice #10

vii. The grinding progress is not only influenced by the ball charge but also by the
material load.

viii. Parallel to the axial sampling and the material load evaluation, the mill internal
is being inspected for coating (on ball charge, liners and diaphragms) The
grading is presented hereafter:
16/10/2007 Page 6 of 20
 Valid Practice #10

o None: Appearance of “metal” as surface

o Light: Appearance of “cement” as surface. Readily cleaned by hand to
reveal “metal” surface.

o Moderate: Appearance of “heavier” cement as surface. Partially cleaned

by hand to reveal “cement/metal” surface.

o Heavy: Appearance of “thick” cement as surface. Difficult to see metal

surface, even after cleaning by hand.

16/10/2007 Page 7 of 20
 Valid Practice #10

…liners condition (cracks, wear):

70mm

Step
Height

…adjustment and condition of the intermediate wall (slope opening):

ix. Finally the mill dimensions are checked by measuring the compartment length
and diameter.
x. After internal mill inspection and material (axial) sampling the maintenance
department will close the doors and turn the mill over to the production
department for restart.

16/10/2007 Page 8 of 20
 Valid Practice #10

1.1.2. Guideline for Empty Grinding Stop

When the mill inspection is finished, after the crash stop, the mill should run for
~30minutes without fresh feed and following should be stopped again (empty
grinding stop). Visual inspection of mill interior along with measurements (free
height, ball charge distribution, etc.) will be carried out. Visual inspection of static
& dynamic separator is also necessary.
i. All associated departments should be informed at least 24hours prior to crash
stopping the mill. The departments involved include Production, Maintenance,
Process and the Quality Control.
ii. The CM must be stopped after running for ~30minutes without fresh feed so
that all material to be ground between the grinding media is discharged from
the mill.
iii. The grinding plant is then stopped with EMERGENCY STOP, while the mill
fan should continue operating in order to cool the mill interior.
iv. The mill will be prepared for inspection according to the respective safety
procedure. Estimated time for preparations is 2hours. It should be noted that
at all time the fuses should be removed from the respective motors and that
nobody should enter without clearance from the Production manager and the
Maintenance Manager.
v. The mill is allowed to cool down. Estimated time for cooling is 3 hours.
vi. In order to calculate the filling degree:
a. the height should be measured along the centerline of the mill from the
top of the balls to the top of mill (minimum 3 locations in 1 st compartment
and 4 locations in 2nd compartment).

b. the chord length should be measured on top of the ball charge level
(minimum 3 locations in 1st compartment and 4 locations in 2nd
compartment).
c. the number of free liner plates should be measured (minimum 3
locations in 1st compartment and 4 locations in 2nd compartment). Make
sure that the first and last plates (% plate) are estimated with as much
accuracy as possible.

16/10/2007 Page 9 of 20
 Valid Practice #10

vii. To check the ball charge quality and composition, ball charge sampling is
performed at the same principle as material axial sampling. Cement bags are
filled with balls after every 1m from the mill inlet to the mill outlet.
viii. After internal mill inspection and material (axial) sampling the maintenance
department will close the doors and turn the mill over to the production
department for restart.

16/10/2007 Page 10 of 20
 Valid Practice #10

2. BALL CEMENT MILL EVALUATION

Sizing of the existing mill equipment is usually are not a factor which can be
optimized. In order to improve mill performance the target should be to follow the
benchmark regarding the production rate and the SEEC. Output and energy
consumption of a cement mill are interdependent and are combined in the kWh/t
of cement produced performance benchmark. The power drawn by the mill main
motor is dependent on the grinding media charge level in the mill rather than the
amount of material being ground in the mill. There is no opportunity to vary this
power drawn during operation of the mill. Whether the productivity is as high as it
could be for the type and fineness of cement being produced is a difficult question
to answer. Comparing performance between milling systems involves complex
peer group analysis.
 The first consideration is the type of cement being produced. Blended
cements have a higher throughput than when grinding pure Portland
cement due to the lower grindability or inherent fineness of the
extender. A cement mill grinding blended cement with granulated slag
additions will have a lower throughput than when grinding pure
Portland cement due to the hardness of the slag.
 The second consideration must be the fineness of the cement
produced on the mill. The kW hours consumed per ton of cement
produced are related to the surface area (cm2/g).
 The third factor that must be taken into consideration is the complexity
of the grinding circuit as well as the respective mill and separator type.
Recognize the overriding importance of quality and focus the process control on
the whole process of cement grinding rather than simply the fineness.
The results of the mill inspection measurements need to be compared with
accepted norms and benchmarks from within the Group or elsewhere in the
cement industry. Certain benchmarks, or rules of thumb, are well established:
 length to diameter ratio should be close to 3.5 for a closed circuit mill
 modern mills operate at 70%-75% of critical speed
 1.5m/s-2.0m/s air velocity through the mill is normal for closed circuit
mill
 1.0m/s-1.5m/s for an open circuit mill
 1st chamber performance
o chamber one occupies 30%-35% of the length of a two chamber
mill
o High volume loading (>30%). 28-30% media volume loading is
normal for a closed circuit mill, exceptionally up to 33%
o Not excessively large feed sizes
o Chamber specific electrical energy consumption 8kWh/t-12kWh/t
o Good Media Grading, 60mm-90mm (all >55mm) , mean size of
75mm-78mm
o Good Liner Step (compared to mill speed), >40mm
o Good diaphragm condition, ~6mm slot width, no gaps, no
blockages
o Select % of 90mm media, usually between 15 and 40% which is
based on feed sizes and crushability (or axial testing).
Automatically calculate remaining charge as equal numbers of
80mm, 70mm and 60mm.
 2nd chamber performance
16/10/2007 Page 11 of 20
 Valid Practice #10

o Absence of “nibs” of clinker or other feed

o Moderate volume loading (25%-35%). Second chamber media
volume loading should be lower than the first chamber
o Mean ball size of 18mm-30mm
o Good Media Classification, if media is up to 60mm
o Minimal Coating on liners and grinding media
o Good diaphragm condition, ~8mm slot width, no gaps, no
blockages
2.1. Relative mill speed calculation
The mill speed should typically range from 70% to 75% of the critical speed.

16/10/2007 Page 12 of 20
 Valid Practice #10

2.2. Axial sampling

At first, any metal parts (balls or ball scrap) must be removed from the
sampled material. The samples are then divided and thus reduced to suitable
quantities for analysis.
It should be noted that:
 The axial sampling results must always be evaluated taking into account
the loading of the mill with material to be ground. Evaluation of the axial
sampling is done according to the following scheme:

 The curves of the coarse particles (above 7mm) in compartment 1

should drop steeply and reach 0 after approx. 50% to 60% of the
effective grinding length of compartment 1. If this is not the case, the ball
charge may be too fine, i.e. the quantity of large balls is not high
enough.
 In a two-compartment mill for cement grinding, the total of the residue
values before the intermediate wall should be between 15% and 25% on
the 0.5mm screen and not more than 5% on the 2mm screen. If this is
not the case, the share of large grinding media is too small and/or the
first grinding compartment contains too much material to be ground. In
that case, a relatively high amount of material is found above the
grinding media and between the grinding media after the emergency
stop.
 For the evaluation of a grinding diagram, it is always necessary to
consider the material load of the mill.
 In the 2nd grinding compartment, the residue values should drop steadily
and the specific surface of the samples should constantly increase.
 With a two-compartment separator mill and a cement fineness (Blaine)
of 3.200cm²/g, the following values should be attained before the
discharge wall:
o Residue on 90μm approximately 15% to 25%
o Residue on 200μm maximally 5%
o Blaine approximately 1.800cm²/g

16/10/2007 Page 13 of 20
 Valid Practice #10

16/10/2007 Page 14 of 20
 Valid Practice #10

2.3. Ball charge calculation

The ball charge is being calculated according to the following formulas, using
the measurements performed during the empty grinding stop. The calculated
ball charge should be compared to the one given by the Plant and should be
confirmed with the power consumption measurements.

16/10/2007 Page 15 of 20
 Valid Practice #10

The ball charge should be confirmed (both periodically i.e. monthly or after every mill
inspection) by the power measurement using the FLS formula:

N=F*D*μ*a*n*g*π/60= F*D*μ*a*n*0,514

a= the arm of gravity in relation to the mill diameter=0.009*(96.7-specific ball charge%*100)

N=power consumption [kW]
F=charge [t]
n=rotational speed of the mill
μ=torque factor
D= effective diameter
Wherever it is necessary, the following tables are used for finding the specific charge &
torque factor:

16/10/2007 Page 16 of 20
 Valid Practice #10

2.4. Ball charge Management

The objective of this manufacturing best practice is to optimize production in a sold-
out market and to reduce the overall kWh/t consumption in a non sold-out market, by
maintaining the optimum ball charges level in all operational mills.
The ideal volume loading is dependent upon the individual plant priorities. The plant
must decide whether their main goal is either to reduce mill power consumption
(kWh/t), or to maximize production (t/h). The recommended volume loading are as
follows:
Recommended Volume Loading
1st Compartment 2nd Compartment
Minimum kWh/t 26 – 28% 28 – 30%
Maximum Production 32 – 34 % 34 – 36%

 It is important to realize that there is higher power consumption (kWh/t) when the
maximum production rate is the first priority.
 Some mills will be limited to lower volume loading due to mill motor power, stress
on mill shell or stress on the gearing.
 Mill inlet trunion size is another possible limiting factor. Ball level in the trunion
should not be higher than 2 to 3 inches.
 With a volume loading above 34% classifying liners often lose their beneficial
effect.

2.4.1. Recommended Ball Charge Sorting Frequency

The minimum frequency for ball charge sorting for mills is recommended as follows:
Ball Sorting
1st Compartment 2nd Compartment
Raw and Coal Mills Every year Every year
Finish Mills Every year Every two years

This is the recommended minimum ball charge sorting. Some plants may require
sorting more often. For mills with very low utilization, the ball sorting may be less
frequent.
2.4.2. Ball Mill Loading
 After sorting add 85 to 90% of ball charge initially. Run ~100 hours before
topping up the ball charge, this allows addition of larger or smaller balls according
to mill inspection and fineness obtained at the partition for the 1st compartment.
 After monthly mill inspection
o Balls should be added to get back to the original ball charge design.
o The average size of added balls will be larger than the average size of the ball
charge design, unless the ball charge design is changed.
o A ball charge sampling should be done if the current ball charge is uncertain,
in order to determine what ball sizes to add.
o An important rule of thumb is that one metric ton of balls will increase the mill
power draw by 10kW. The required mill kilowatts must be available before
adding the balls to the mill.
 Do not mix different types and metallurgical compositions of balls together, i.e.:
o forged steel and chromium
o low chrome and high chrome balls in the 1st compartment
o balls from different supplier, unless proven effective by extensive testing and
quality control
Exception: Used first compartment balls could be recycled in the second
compartment, if they have the appropriate size (Not mandatory)
 Do not add balls that are wet or snow covered to cement mills.
16/10/2007 Page 17 of 20
 Valid Practice #10

 Add a layer of material to the first compartment only before adding balls to an
empty mill, in order to protect the shell liners.
 Never exceed mill motor power.
2.4.3. Ball charge record keeping
The following items should be kept as records in plants for proper ball charge
management:
 Monthly mill ball charge inspection sheet.
 Ball additions.
 Ball charge sorting results.
 Ball charge design.
 Any unusual occurrence or problem that affected mill performance.
 Dates of liner and partition changes

16/10/2007 Page 18 of 20
 Valid Practice #10

2.4.4. Ball sorting

It is usual, that after some time the ball charge has to be sorted in order to replace
the worn parts and establish ball charge composition according to the axial sampling
results as well as the granulometry. This can be done either manually or using
specific equipment as following:

16/10/2007 Page 19 of 20
 Valid Practice #10

2.5. Conclusions
Depending on the findings, the most possible actions may include:
 Once per month, the chambers should be charged with the required grinding
media. The required grinding media can be calculated by 3 ways:
 Mill motor kW trend
 Monthly production combined with the statistical grinding media
consumption (g/t)
 Check of the grinding media level at the two chambers
 Increase the ball charge according to the Plant policy (either for minimum
energy consumption or for the maximum production capacity), based on the
installed motor power and the typical reference values.
 Sorting the ball charge of the both chambers. Once per year, sorting of the
ball charge is required.
 Based on axial sampling the ball charge composition or the location of the
intermediate diaphragm could be changed.
 The specific electrical energy consumption is compared to the typical
references.
 The separator performance could imply the presence of a mechanical failure.
 The gas measurements should be used for identifying the source of false air
induction.
 The impact flow-meter of separator rejects needs to be calibrated periodically.
 The folafon indication should be calibrated periodically.
 Based on the main reason for stoppages the run factor is being evaluated and
justified.

16/10/2007 Page 20 of 20